Search Results for author: Junchi Yan

Found 207 papers, 97 papers with code

Pre-training Entity Relation Encoder with Intra-span and Inter-span Information

no code implementations EMNLP 2020 Yijun Wang, Changzhi Sun, Yuanbin Wu, Junchi Yan, Peng Gao, Guotong Xie

In particular, a span encoder is trained to recover a random shuffling of tokens in a span, and a span pair encoder is trained to predict positive pairs that are from the same sentences and negative pairs that are from different sentences using contrastive loss.

Relation Relation Extraction +1

Rethinking the Defocus Blur Detection Problem and A Real-Time Deep DBD Model

no code implementations ECCV 2020 Ning Zhang, Junchi Yan

In this work, we propose novel perspectives on the DBD problem and design convenient approach to build a real-time cost-effective DBD model.

Data Augmentation Defocus Blur Detection

AutoMix: Mixup Networks for Sample Interpolation via Cooperative Barycenter Learning

no code implementations ECCV 2020 Jianchao Zhu, Liangliang Shi, Junchi Yan, Hongyuan Zha

This paper proposes new ways of sample mixing by thinking of the process as generation of barycenter in a metric space for data augmentation.

Data Augmentation

Discriminative Partial Domain Adversarial Network

no code implementations ECCV 2020 Jian Hu, Hongya Tuo, Chao Wang, Lingfeng Qiao, Haowen Zhong, Junchi Yan, Zhongliang Jing, Henry Leung

Previous methods typically match the whole source domain to target domain, which causes negative transfer due to the source-negative classes in source domain that does not exist in target domain.

Partial Domain Adaptation Transfer Learning

AMP: Autoregressive Motion Prediction Revisited with Next Token Prediction for Autonomous Driving

no code implementations20 Mar 2024 Xiaosong Jia, Shaoshuai Shi, Zijun Chen, Li Jiang, Wenlong Liao, Tao He, Junchi Yan

As an essential task in autonomous driving (AD), motion prediction aims to predict the future states of surround objects for navigation.

Motion Forecasting motion prediction +1

Boosting Order-Preserving and Transferability for Neural Architecture Search: a Joint Architecture Refined Search and Fine-tuning Approach

1 code implementation18 Mar 2024 Beichen Zhang, Xiaoxing Wang, Xiaohan Qin, Junchi Yan

In this work, we analyze the order-preserving ability on the whole search space (global) and a sub-space of top architectures (local), and empirically show that the local order-preserving for current two-stage NAS methods still need to be improved.

Neural Architecture Search

Boundary Matters: A Bi-Level Active Finetuning Framework

no code implementations15 Mar 2024 Han Lu, Yichen Xie, Xiaokang Yang, Junchi Yan

In this paper, we propose a Bi-Level Active Finetuning framework to select the samples for annotation in one shot, which includes two stages: core sample selection for diversity, and boundary sample selection for uncertainty.

Active Learning Denoising

Exploring Safety Generalization Challenges of Large Language Models via Code

no code implementations12 Mar 2024 Qibing Ren, Chang Gao, Jing Shao, Junchi Yan, Xin Tan, Yu Qiao, Wai Lam, Lizhuang Ma

The rapid advancement of Large Language Models (LLMs) has brought about remarkable capabilities in natural language processing but also raised concerns about their potential misuse.

Rethinking Classifier Re-Training in Long-Tailed Recognition: A Simple Logits Retargeting Approach

no code implementations1 Mar 2024 Han Lu, Siyu Sun, Yichen Xie, Liqing Zhang, Xiaokang Yang, Junchi Yan

In the long-tailed recognition field, the Decoupled Training paradigm has demonstrated remarkable capabilities among various methods.

Representation Learning

Fast and Interpretable 2D Homography Decomposition: Similarity-Kernel-Similarity and Affine-Core-Affine Transformations

1 code implementation28 Feb 2024 Shen Cai, Zhanhao Wu, Lingxi Guo, Jiachun Wang, Siyu Zhang, Junchi Yan, Shuhan Shen

Under the minimal $4$-point configuration, the first and the last similarity transformations in SKS are computed by two anchor points on target and source planes, respectively.

Computational Efficiency

Not All Experts are Equal: Efficient Expert Pruning and Skipping for Mixture-of-Experts Large Language Models

1 code implementation22 Feb 2024 Xudong Lu, Qi Liu, Yuhui Xu, Aojun Zhou, Siyuan Huang, Bo Zhang, Junchi Yan, Hongsheng Li

Specifically, we propose, for the first time to our best knowledge, post-training approaches for task-agnostic and task-specific expert pruning and skipping of MoE LLMs, tailored to improve deployment efficiency while maintaining model performance across a wide range of tasks.

Graph Out-of-Distribution Generalization via Causal Intervention

1 code implementation18 Feb 2024 Qitian Wu, Fan Nie, Chenxiao Yang, TianYi Bao, Junchi Yan

In this paper, we adopt a bottom-up data-generative perspective and reveal a key observation through causal analysis: the crux of GNNs' failure in OOD generalization lies in the latent confounding bias from the environment.

Causal Inference Out-of-Distribution Generalization

GSINA: Improving Subgraph Extraction for Graph Invariant Learning via Graph Sinkhorn Attention

1 code implementation11 Feb 2024 Fangyu Ding, Haiyang Wang, Zhixuan Chu, Tianming Li, Zhaoping Hu, Junchi Yan

Many recent endeavors of GIL focus on extracting the invariant subgraph from the input graph for prediction as a regularization strategy to improve the generalization performance of graph learning.

Graph Attention Graph Learning

Cross-Task Linearity Emerges in the Pretraining-Finetuning Paradigm

no code implementations6 Feb 2024 Zhanpeng Zhou, Zijun Chen, Yilan Chen, Bo Zhang, Junchi Yan

The pretraining-finetuning paradigm has become the prevailing trend in modern deep learning.

Poisson Process for Bayesian Optimization

no code implementations5 Feb 2024 Xiaoxing Wang, Jiaxing Li, Chao Xue, Wei Liu, Weifeng Liu, Xiaokang Yang, Junchi Yan, DaCheng Tao

BayesianOptimization(BO) is a sample-efficient black-box optimizer, and extensive methods have been proposed to build the absolute function response of the black-box function through a probabilistic surrogate model, including Tree-structured Parzen Estimator (TPE), random forest (SMAC), and Gaussian process (GP).

Bayesian Optimization Hyperparameter Optimization +2

ViTree: Single-path Neural Tree for Step-wise Interpretable Fine-grained Visual Categorization

no code implementations30 Jan 2024 Danning Lao, Qi Liu, Jiazi Bu, Junchi Yan, Wei Shen

As computer vision continues to advance and finds widespread applications across various domains, the need for interpretability in deep learning models becomes paramount.

Decision Making Fine-Grained Visual Categorization

Continuous-Multiple Image Outpainting in One-Step via Positional Query and A Diffusion-based Approach

1 code implementation28 Jan 2024 Shaofeng Zhang, Jinfa Huang, Qiang Zhou, Zhibin Wang, Fan Wang, Jiebo Luo, Junchi Yan

At inference, we generate images with arbitrary expansion multiples by inputting an anchor image and its corresponding positional embeddings.

Image Outpainting

Knowledge Guided Entity-aware Video Captioning and A Basketball Benchmark

no code implementations25 Jan 2024 Zeyu Xi, Ge Shi, Xuefen Li, Junchi Yan, Zun Li, Lifang Wu, Zilin Liu, Liang Wang

We develop a knowledge guided entity-aware video captioning network (KEANet) based on a candidate player list in encoder-decoder form for basketball live text broadcast.

Video Captioning

Double-Bounded Optimal Transport for Advanced Clustering and Classification

no code implementations21 Jan 2024 Liangliang Shi, Zhaoqi Shen, Junchi Yan

Even with vanilla Softmax trained features, our extensive experimental results show that our method can achieve good results with our improved inference scheme in the testing stage.

Clustering

Machine Learning Insides OptVerse AI Solver: Design Principles and Applications

no code implementations11 Jan 2024 Xijun Li, Fangzhou Zhu, Hui-Ling Zhen, Weilin Luo, Meng Lu, Yimin Huang, Zhenan Fan, Zirui Zhou, Yufei Kuang, Zhihai Wang, Zijie Geng, Yang Li, Haoyang Liu, Zhiwu An, Muming Yang, Jianshu Li, Jie Wang, Junchi Yan, Defeng Sun, Tao Zhong, Yong Zhang, Jia Zeng, Mingxuan Yuan, Jianye Hao, Jun Yao, Kun Mao

To this end, we present a comprehensive study on the integration of machine learning (ML) techniques into Huawei Cloud's OptVerse AI Solver, which aims to mitigate the scarcity of real-world mathematical programming instances, and to surpass the capabilities of traditional optimization techniques.

Decision Making Management

LLM4EDA: Emerging Progress in Large Language Models for Electronic Design Automation

1 code implementation28 Dec 2023 RuiZhe Zhong, Xingbo Du, Shixiong Kai, Zhentao Tang, Siyuan Xu, Hui-Ling Zhen, Jianye Hao, Qiang Xu, Mingxuan Yuan, Junchi Yan

Since circuit can be represented with HDL in a textual format, it is reasonable to question whether LLMs can be leveraged in the EDA field to achieve fully automated chip design and generate circuits with improved power, performance, and area (PPA).

Answer Generation Chatbot

LaneSegNet: Map Learning with Lane Segment Perception for Autonomous Driving

1 code implementation26 Dec 2023 Tianyu Li, Peijin Jia, Bangjun Wang, Li Chen, Kun Jiang, Junchi Yan, Hongyang Li

A map, as crucial information for downstream applications of an autonomous driving system, is usually represented in lanelines or centerlines.

Autonomous Driving

Open-sourced Data Ecosystem in Autonomous Driving: the Present and Future

2 code implementations6 Dec 2023 Hongyang Li, Yang Li, Huijie Wang, Jia Zeng, Huilin Xu, Pinlong Cai, Li Chen, Junchi Yan, Feng Xu, Lu Xiong, Jingdong Wang, Futang Zhu, Chunjing Xu, Tiancai Wang, Fei Xia, Beipeng Mu, Zhihui Peng, Dahua Lin, Yu Qiao

With the continuous maturation and application of autonomous driving technology, a systematic examination of open-source autonomous driving datasets becomes instrumental in fostering the robust evolution of the industry ecosystem.

Autonomous Driving

Rethinking Urban Mobility Prediction: A Super-Multivariate Time Series Forecasting Approach

1 code implementation4 Dec 2023 Jinguo Cheng, Ke Li, Yuxuan Liang, Lijun Sun, Junchi Yan, Yuankai Wu

To address this challenge, we present the Super-Multivariate Urban Mobility Transformer (SUMformer), which utilizes a specially designed attention mechanism to calculate temporal and cross-variable correlations and reduce computational costs stemming from a large number of time series.

Multivariate Time Series Forecasting Time Series +1

Unified Batch Normalization: Identifying and Alleviating the Feature Condensation in Batch Normalization and a Unified Framework

no code implementations27 Nov 2023 Shaobo Wang, Xiangdong Zhang, Dongrui Liu, Junchi Yan

In this work, we critically examine BN from a feature perspective, identifying feature condensation during BN as a detrimental factor to test performance.

Instance Segmentation object-detection +2

GMTR: Graph Matching Transformers

1 code implementation14 Nov 2023 Jinpei Guo, Shaofeng Zhang, Runzhong Wang, Chang Liu, Junchi Yan

Meanwhile, on Pascal VOC, QueryTrans improves the accuracy of NGMv2 from $80. 1\%$ to $\mathbf{83. 3\%}$, and BBGM from $79. 0\%$ to $\mathbf{84. 5\%}$.

 Ranked #1 on Graph Matching on PASCAL VOC (matching accuracy metric)

Graph Attention Graph Matching +2

Rethinking and Benchmarking Predict-then-Optimize Paradigm for Combinatorial Optimization Problems

no code implementations13 Nov 2023 Haoyu Geng, Hang Ruan, Runzhong Wang, Yang Li, Yang Wang, Lei Chen, Junchi Yan

Numerous web applications rely on solving combinatorial optimization problems, such as energy cost-aware scheduling, budget allocation on web advertising, and graph matching on social networks.

Benchmarking Combinatorial Optimization +3

Rethinking Cross-Domain Sequential Recommendation under Open-World Assumptions

1 code implementation8 Nov 2023 Wujiang Xu, Qitian Wu, Runzhong Wang, Mingming Ha, Qiongxu Ma, Linxun Chen, Bing Han, Junchi Yan

To address these challenges under open-world assumptions, we design an \textbf{A}daptive \textbf{M}ulti-\textbf{I}nterest \textbf{D}ebiasing framework for cross-domain sequential recommendation (\textbf{AMID}), which consists of a multi-interest information module (\textbf{MIM}) and a doubly robust estimator (\textbf{DRE}).

Sequential Recommendation

RigLSTM: Recurrent Independent Grid LSTM for Generalizable Sequence Learning

no code implementations3 Nov 2023 Ziyu Wang, Wenhao Jiang, Zixuan Zhang, Wei Tang, Junchi Yan

Sequential processes in real-world often carry a combination of simple subsystems that interact with each other in certain forms.

feature selection

LLM4Drive: A Survey of Large Language Models for Autonomous Driving

1 code implementation2 Nov 2023 Zhenjie Yang, Xiaosong Jia, Hongyang Li, Junchi Yan

Recently, large language models (LLMs) have demonstrated abilities including understanding context, logical reasoning, and generating answers.

Autonomous Driving Few-Shot Learning +1

On the Evaluation and Refinement of Vision-Language Instruction Tuning Datasets

no code implementations10 Oct 2023 Ning Liao, Shaofeng Zhang, Renqiu Xia, Min Cao, Yu Qiao, Junchi Yan

Instead of evaluating the models directly, in this paper, we try to evaluate the Vision-Language Instruction-Tuning (VLIT) datasets.

Benchmarking

Advective Diffusion Transformers for Topological Generalization in Graph Learning

no code implementations10 Oct 2023 Qitian Wu, Chenxiao Yang, Kaipeng Zeng, Fan Nie, Michael Bronstein, Junchi Yan

Graph diffusion equations are intimately related to graph neural networks (GNNs) and have recently attracted attention as a principled framework for analyzing GNN dynamics, formalizing their expressive power, and justifying architectural choices.

Graph Learning

Going Beyond Neural Network Feature Similarity: The Network Feature Complexity and Its Interpretation Using Category Theory

no code implementations10 Oct 2023 Yiting Chen, Zhanpeng Zhou, Junchi Yan

In this paper, we expand the concept of equivalent feature and provide the definition of what we call functionally equivalent features.

How Graph Neural Networks Learn: Lessons from Training Dynamics

no code implementations8 Oct 2023 Chenxiao Yang, Qitian Wu, David Wipf, Ruoyu Sun, Junchi Yan

A long-standing goal in deep learning has been to characterize the learning behavior of black-box models in a more interpretable manner.

Inductive Bias

StructChart: Perception, Structuring, Reasoning for Visual Chart Understanding

1 code implementation20 Sep 2023 Renqiu Xia, Bo Zhang, Haoyang Peng, Hancheng Ye, Xiangchao Yan, Peng Ye, Botian Shi, Yu Qiao, Junchi Yan

Charts are common in literature across different scientific fields, conveying rich information easily accessible to readers.

Ranked #17 on Chart Question Answering on ChartQA (using extra training data)

Chart Question Answering Language Modelling +2

SPOT: Scalable 3D Pre-training via Occupancy Prediction for Autonomous Driving

1 code implementation19 Sep 2023 Xiangchao Yan, Runjian Chen, Bo Zhang, Jiakang Yuan, Xinyu Cai, Botian Shi, Wenqi Shao, Junchi Yan, Ping Luo, Yu Qiao

Our contributions are threefold: (1) Occupancy prediction is shown to be promising for learning general representations, which is demonstrated by extensive experiments on plenty of datasets and tasks.

3D Object Detection Autonomous Driving +3

ReSimAD: Zero-Shot 3D Domain Transfer for Autonomous Driving with Source Reconstruction and Target Simulation

2 code implementations11 Sep 2023 Bo Zhang, Xinyu Cai, Jiakang Yuan, Donglin Yang, Jianfei Guo, Xiangchao Yan, Renqiu Xia, Botian Shi, Min Dou, Tao Chen, Si Liu, Junchi Yan, Yu Qiao

Domain shifts such as sensor type changes and geographical situation variations are prevalent in Autonomous Driving (AD), which poses a challenge since AD model relying on the previous domain knowledge can be hardly directly deployed to a new domain without additional costs.

Autonomous Driving Domain Generalization

DriveAdapter: Breaking the Coupling Barrier of Perception and Planning in End-to-End Autonomous Driving

1 code implementation ICCV 2023 Xiaosong Jia, Yulu Gao, Li Chen, Junchi Yan, Patrick Langechuan Liu, Hongyang Li

We find that even equipped with a SOTA perception model, directly letting the student model learn the required inputs of the teacher model leads to poor driving performance, which comes from the large distribution gap between predicted privileged inputs and the ground-truth.

Autonomous Driving CARLA longest6

Patch-Level Contrasting without Patch Correspondence for Accurate and Dense Contrastive Representation Learning

no code implementations23 Jun 2023 Shaofeng Zhang, Feng Zhu, Rui Zhao, Junchi Yan

On classification tasks, for ViT-S, ADCLR achieves 77. 5% top-1 accuracy on ImageNet with linear probing, outperforming our baseline (DINO) without our devised techniques as plug-in, by 0. 5%.

Instance Segmentation object-detection +4

GraphGLOW: Universal and Generalizable Structure Learning for Graph Neural Networks

1 code implementation20 Jun 2023 Wentao Zhao, Qitian Wu, Chenxiao Yang, Junchi Yan

Graph structure learning is a well-established problem that aims at optimizing graph structures adaptive to specific graph datasets to help message passing neural networks (i. e., GNNs) to yield effective and robust node embeddings.

Graph structure learning

NodeFormer: A Scalable Graph Structure Learning Transformer for Node Classification

1 code implementation14 Jun 2023 Qitian Wu, Wentao Zhao, Zenan Li, David Wipf, Junchi Yan

In this paper, we introduce a novel all-pair message passing scheme for efficiently propagating node signals between arbitrary nodes, as an important building block for a pioneering Transformer-style network for node classification on large graphs, dubbed as \textsc{NodeFormer}.

Graph structure learning Image Classification

Geometric-aware Pretraining for Vision-centric 3D Object Detection

1 code implementation6 Apr 2023 Linyan Huang, Huijie Wang, Jia Zeng, Shengchuan Zhang, Liujuan Cao, Junchi Yan, Hongyang Li

We also conduct experiments on various image backbones and view transformations to validate the efficacy of our approach.

3D Object Detection Autonomous Driving +2

Active Finetuning: Exploiting Annotation Budget in the Pretraining-Finetuning Paradigm

1 code implementation CVPR 2023 Yichen Xie, Han Lu, Junchi Yan, Xiaokang Yang, Masayoshi Tomizuka, Wei Zhan

We propose a novel method called ActiveFT for active finetuning task to select a subset of data distributing similarly with the entire unlabeled pool and maintaining enough diversity by optimizing a parametric model in the continuous space.

Image Classification Semantic Segmentation

EasyDGL: Encode, Train and Interpret for Continuous-time Dynamic Graph Learning

1 code implementation22 Mar 2023 Chao Chen, Haoyu Geng, Nianzu Yang, Xiaokang Yang, Junchi Yan

Dynamic graphs arise in various real-world applications, and it is often welcomed to model the dynamics directly in continuous time domain for its flexibility.

Dynamic Link Prediction Dynamic Node Classification +4

Rethinking Visual Prompt Learning as Masked Visual Token Modeling

no code implementations9 Mar 2023 Ning Liao, Bowen Shi, Xiaopeng Zhang, Min Cao, Junchi Yan, Qi Tian

To explore prompt learning on the generative pre-trained visual model, as well as keeping the task consistency, we propose Visual Prompt learning as masked visual Token Modeling (VPTM) to transform the downstream visual classification into the pre-trained masked visual token prediction.

ARS-DETR: Aspect Ratio Sensitive Oriented Object Detection with Transformer

1 code implementation9 Mar 2023 Ying Zeng, Xue Yang, Qingyun Li, Yushi Chen, Junchi Yan

Existing oriented object detection methods commonly use metric AP$_{50}$ to measure the performance of the model.

Object object-detection +2

M-Tuning: Prompt Tuning with Mitigated Label Bias in Open-Set Scenarios

no code implementations9 Mar 2023 Ning Liao, Xiaopeng Zhang, Min Cao, Junchi Yan, Qi Tian

In realistic open-set scenarios where labels of a part of testing data are totally unknown, when vision-language (VL) prompt learning methods encounter inputs related to unknown classes (i. e., not seen during training), they always predict them as one of the training classes.

Open Set Learning

Graph Signal Sampling for Inductive One-Bit Matrix Completion: a Closed-form Solution

1 code implementation8 Feb 2023 Chao Chen, Haoyu Geng, Gang Zeng, Zhaobing Han, Hua Chai, Xiaokang Yang, Junchi Yan

Inductive one-bit matrix completion is motivated by modern applications such as recommender systems, where new users would appear at test stage with the ratings consisting of only ones and no zeros.

Matrix Completion Recommendation Systems

Energy-based Out-of-Distribution Detection for Graph Neural Networks

1 code implementation6 Feb 2023 Qitian Wu, Yiting Chen, Chenxiao Yang, Junchi Yan

This paves a way for a simple, powerful and efficient OOD detection model for GNN-based learning on graphs, which we call GNNSafe.

Out-of-Distribution Detection Out of Distribution (OOD) Detection

HardSATGEN: Understanding the Difficulty of Hard SAT Formula Generation and A Strong Structure-Hardness-Aware Baseline

1 code implementation4 Feb 2023 Yang Li, Xinyan Chen, Wenxuan Guo, Xijun Li, Wanqian Luo, Junhua Huang, Hui-Ling Zhen, Mingxuan Yuan, Junchi Yan

On top of the observations that industrial formulae exhibit clear community structure and oversplit substructures lead to the difficulty in semantic formation of logical structures, we propose HardSATGEN, which introduces a fine-grained control mechanism to the neural split-merge paradigm for SAT formula generation to better recover the structural and computational properties of the industrial benchmarks.

DIFFormer: Scalable (Graph) Transformers Induced by Energy Constrained Diffusion

1 code implementation23 Jan 2023 Qitian Wu, Chenxiao Yang, Wentao Zhao, Yixuan He, David Wipf, Junchi Yan

Real-world data generation often involves complex inter-dependencies among instances, violating the IID-data hypothesis of standard learning paradigms and posing a challenge for uncovering the geometric structures for learning desired instance representations.

Image-text Classification Node Classification +2

Policy Pre-training for Autonomous Driving via Self-supervised Geometric Modeling

1 code implementation3 Jan 2023 Penghao Wu, Li Chen, Hongyang Li, Xiaosong Jia, Junchi Yan, Yu Qiao

Witnessing the impressive achievements of pre-training techniques on large-scale data in the field of computer vision and natural language processing, we wonder whether this idea could be adapted in a grab-and-go spirit, and mitigate the sample inefficiency problem for visuomotor driving.

Autonomous Driving Decision Making

Deep Learning of Partial Graph Matching via Differentiable Top-K

1 code implementation CVPR 2023 Runzhong Wang, Ziao Guo, Shaofei Jiang, Xiaokang Yang, Junchi Yan

Graph matching (GM) aims at discovering node matching between graphs, by maximizing the node- and edge-wise affinities between the matched elements.

 Ranked #1 on Graph Matching on Willow Object Class (F1 score metric)

Graph Matching Stereo Matching

Distilling Focal Knowledge From Imperfect Expert for 3D Object Detection

no code implementations CVPR 2023 Jia Zeng, Li Chen, Hanming Deng, Lewei Lu, Junchi Yan, Yu Qiao, Hongyang Li

Specifically, a set of queries are leveraged to locate the instance-level areas for masked feature generation, to intensify feature representation ability in these areas.

3D Object Detection Knowledge Distillation +2

Graph Neural Networks are Inherently Good Generalizers: Insights by Bridging GNNs and MLPs

1 code implementation18 Dec 2022 Chenxiao Yang, Qitian Wu, Jiahua Wang, Junchi Yan

Graph neural networks (GNNs), as the de-facto model class for representation learning on graphs, are built upon the multi-layer perceptrons (MLP) architecture with additional message passing layers to allow features to flow across nodes.

Representation Learning

Localized Contrastive Learning on Graphs

no code implementations8 Dec 2022 Hengrui Zhang, Qitian Wu, Yu Wang, Shaofeng Zhang, Junchi Yan, Philip S. Yu

Contrastive learning methods based on InfoNCE loss are popular in node representation learning tasks on graph-structured data.

Contrastive Learning Data Augmentation +1

Leveraging Angular Information Between Feature and Classifier for Long-tailed Learning: A Prediction Reformulation Approach

no code implementations3 Dec 2022 Haoxuan Wang, Junchi Yan

Deep neural networks still struggle on long-tailed image datasets, and one of the reasons is that the imbalance of training data across categories leads to the imbalance of trained model parameters.

Long-tail Learning

Rethinking and Improving Robustness of Convolutional Neural Networks: a Shapley Value-based Approach in Frequency Domain

1 code implementation NIPS 2022 Yiting Chen, Qibing Ren, Junchi Yan

In this work, we introduce Shapley value, a metric of cooperative game theory, into the frequency domain and propose to quantify the positive (negative) impact of every frequency component of data on CNNs.

Adversarial Attack Adversarial Robustness +3

Geometric Knowledge Distillation: Topology Compression for Graph Neural Networks

2 code implementations24 Oct 2022 Chenxiao Yang, Qitian Wu, Junchi Yan

We study a new paradigm of knowledge transfer that aims at encoding graph topological information into graph neural networks (GNNs) by distilling knowledge from a teacher GNN model trained on a complete graph to a student GNN model operating on a smaller or sparser graph.

Knowledge Distillation Transfer Learning

Towards Out-of-Distribution Sequential Event Prediction: A Causal Treatment

1 code implementation24 Oct 2022 Chenxiao Yang, Qitian Wu, Qingsong Wen, Zhiqiang Zhou, Liang Sun, Junchi Yan

The goal of sequential event prediction is to estimate the next event based on a sequence of historical events, with applications to sequential recommendation, user behavior analysis and clinical treatment.

Sequential Recommendation Variational Inference

End-to-End Context-Aided Unicity Matching for Person Re-identification

no code implementations20 Oct 2022 Min Cao, Cong Ding, Chen Chen, Junchi Yan, Qi Tian

Based on a natural assumption that images belonging to the same person identity should not match with images belonging to multiple different person identities across views, called the unicity of person matching on the identity level, we propose an end-to-end person unicity matching architecture for learning and refining the person matching relations.

Graph Matching Person Re-Identification

Learning Universe Model for Partial Matching Networks over Multiple Graphs

no code implementations19 Oct 2022 Zetian Jiang, Jiaxin Lu, Tianzhe Wang, Junchi Yan

We consider the general setting for partial matching of two or multiple graphs, in the sense that not necessarily all the nodes in one graph can find their correspondences in another graph and vice versa.

Graph Matching Metric Learning +1

H2RBox: Horizontal Box Annotation is All You Need for Oriented Object Detection

2 code implementations13 Oct 2022 Xue Yang, Gefan Zhang, Wentong Li, Xuehui Wang, Yue Zhou, Junchi Yan

Oriented object detection emerges in many applications from aerial images to autonomous driving, while many existing detection benchmarks are annotated with horizontal bounding box only which is also less costive than fine-grained rotated box, leading to a gap between the readily available training corpus and the rising demand for oriented object detection.

Autonomous Driving Box-supervised Instance Segmentation +6

Detecting Rotated Objects as Gaussian Distributions and Its 3-D Generalization

1 code implementation22 Sep 2022 Xue Yang, Gefan Zhang, Xiaojiang Yang, Yue Zhou, Wentao Wang, Jin Tang, Tao He, Junchi Yan

Existing detection methods commonly use a parameterized bounding box (BBox) to model and detect (horizontal) objects and an additional rotation angle parameter is used for rotated objects.

regression

ST-P3: End-to-end Vision-based Autonomous Driving via Spatial-Temporal Feature Learning

1 code implementation15 Jul 2022 Shengchao Hu, Li Chen, Penghao Wu, Hongyang Li, Junchi Yan, DaCheng Tao

In particular, we propose a spatial-temporal feature learning scheme towards a set of more representative features for perception, prediction and planning tasks simultaneously, which is called ST-P3.

Ranked #7 on Bird's-Eye View Semantic Segmentation on nuScenes (IoU ped - 224x480 - Vis filter. - 100x100 at 0.5 metric)

Autonomous Driving Bird's-Eye View Semantic Segmentation +1

Level 2 Autonomous Driving on a Single Device: Diving into the Devils of Openpilot

no code implementations16 Jun 2022 Li Chen, Tutian Tang, Zhitian Cai, Yang Li, Penghao Wu, Hongyang Li, Jianping Shi, Junchi Yan, Yu Qiao

Equipped with a wide span of sensors, predominant autonomous driving solutions are becoming more modular-oriented for safe system design.

Autonomous Driving

Learning Unbiased Transferability for Domain Adaptation by Uncertainty Modeling

1 code implementation2 Jun 2022 Jian Hu, Haowen Zhong, Junchi Yan, Shaogang Gong, Guile Wu, Fei Yang

However, due to the significant imbalance between the amount of annotated data in the source and target domains, usually only the target distribution is aligned to the source domain, leading to adapting unnecessary source specific knowledge to the target domain, i. e., biased domain adaptation.

Domain Adaptation Pseudo Label +1

CEP3: Community Event Prediction with Neural Point Process on Graph

no code implementations21 May 2022 Xuhong Wang, Sirui Chen, Yixuan He, Minjie Wang, Quan Gan, Yupu Yang, Junchi Yan

Many real world applications can be formulated as event forecasting on Continuous Time Dynamic Graphs (CTDGs) where the occurrence of a timed event between two entities is represented as an edge along with its occurrence timestamp in the graphs. However, most previous works approach the problem in compromised settings, either formulating it as a link prediction task on the graph given the event time or a time prediction problem given which event will happen next.

Link Prediction

HDGT: Heterogeneous Driving Graph Transformer for Multi-Agent Trajectory Prediction via Scene Encoding

1 code implementation30 Apr 2022 Xiaosong Jia, Penghao Wu, Li Chen, Yu Liu, Hongyang Li, Junchi Yan

Based on these observations, we propose Heterogeneous Driving Graph Transformer (HDGT), a backbone modelling the driving scene as a heterogeneous graph with different types of nodes and edges.

Autonomous Driving graph construction +2

MMRotate: A Rotated Object Detection Benchmark using PyTorch

1 code implementation28 Apr 2022 Yue Zhou, Xue Yang, Gefan Zhang, Jiabao Wang, Yanyi Liu, Liping Hou, Xue Jiang, Xingzhao Liu, Junchi Yan, Chengqi Lyu, Wenwei Zhang, Kai Chen

We present an open-source toolbox, named MMRotate, which provides a coherent algorithm framework of training, inferring, and evaluation for the popular rotated object detection algorithm based on deep learning.

Object object-detection +1

MHSCNet: A Multimodal Hierarchical Shot-aware Convolutional Network for Video Summarization

1 code implementation18 Apr 2022 Wujiang Xu, Runzhong Wang, Xiaobo Guo, Shaoshuai Li, Qiongxu Ma, Yunan Zhao, Sheng Guo, Zhenfeng Zhu, Junchi Yan

However, the optimal video summaries need to reflect the most valuable keyframe with its own information, and one with semantic power of the whole content.

Video Summarization

Learning Self-Modulating Attention in Continuous Time Space with Applications to Sequential Recommendation

1 code implementation30 Mar 2022 Chao Chen, Haoyu Geng, Nianzu Yang, Junchi Yan, Daiyue Xue, Jianping Yu, Xiaokang Yang

User interests are usually dynamic in the real world, which poses both theoretical and practical challenges for learning accurate preferences from rich behavior data.

Dynamic Link Prediction Sequential Recommendation

On Understanding and Mitigating the Dimensional Collapse of Graph Contrastive Learning: a Non-Maximum Removal Approach

no code implementations24 Mar 2022 Jiawei Sun, Ruoxin Chen, Jie Li, Chentao Wu, Yue Ding, Junchi Yan

Graph Contrastive Learning (GCL) has shown promising performance in graph representation learning (GRL) without the supervision of manual annotations.

Contrastive Learning Graph Classification +1

PersFormer: 3D Lane Detection via Perspective Transformer and the OpenLane Benchmark

2 code implementations21 Mar 2022 Li Chen, Chonghao Sima, Yang Li, Zehan Zheng, Jiajie Xu, Xiangwei Geng, Hongyang Li, Conghui He, Jianping Shi, Yu Qiao, Junchi Yan

Methods for 3D lane detection have been recently proposed to address the issue of inaccurate lane layouts in many autonomous driving scenarios (uphill/downhill, bump, etc.).

3D Lane Detection Autonomous Driving +1

EAutoDet: Efficient Architecture Search for Object Detection

no code implementations21 Mar 2022 Xiaoxing Wang, Jiale Lin, Junchi Yan, Juanping Zhao, Xiaokang Yang

In contrast, this paper introduces an efficient framework, named EAutoDet, that can discover practical backbone and FPN architectures for object detection in 1. 4 GPU-days.

Ranked #30 on Object Detection In Aerial Images on DOTA (using extra training data)

Object object-detection +1

Object Localization under Single Coarse Point Supervision

2 code implementations CVPR 2022 Xuehui Yu, Pengfei Chen, Di wu, Najmul Hassan, Guorong Li, Junchi Yan, Humphrey Shi, Qixiang Ye, Zhenjun Han

In this study, we propose a POL method using coarse point annotations, relaxing the supervision signals from accurate key points to freely spotted points.

Multiple Instance Learning Object +1

Machine Learning Methods in Solving the Boolean Satisfiability Problem

no code implementations2 Mar 2022 Wenxuan Guo, Junchi Yan, Hui-Ling Zhen, Xijun Li, Mingxuan Yuan, Yaohui Jin

This paper reviews the recent literature on solving the Boolean satisfiability problem (SAT), an archetypal NP-complete problem, with the help of machine learning techniques.

BIG-bench Machine Learning

Learning Neural Hamiltonian Dynamics: A Methodological Overview

1 code implementation28 Feb 2022 Zhijie Chen, Mingquan Feng, Junchi Yan, Hongyuan Zha

The past few years have witnessed an increased interest in learning Hamiltonian dynamics in deep learning frameworks.

Inductive Bias

From Quantum Graph Computing to Quantum Graph Learning: A Survey

no code implementations19 Feb 2022 Yehui Tang, Junchi Yan, Hancock Edwin

Quantum computing (QC) is a new computational paradigm whose foundations relate to quantum physics.

Graph Learning

Molecule Generation for Drug Design: a Graph Learning Perspective

no code implementations18 Feb 2022 Nianzu Yang, Huaijin Wu, Kaipeng Zeng, Yang Li, Junchi Yan

Machine learning, particularly graph learning, is gaining increasing recognition for its transformative impact across various fields.

Graph Learning

Transformers in Time Series: A Survey

10 code implementations15 Feb 2022 Qingsong Wen, Tian Zhou, Chaoli Zhang, Weiqi Chen, Ziqing Ma, Junchi Yan, Liang Sun

From the perspective of network structure, we summarize the adaptations and modifications that have been made to Transformers in order to accommodate the challenges in time series analysis.

Anomaly Detection Time Series +1

Handling Distribution Shifts on Graphs: An Invariance Perspective

2 code implementations ICLR 2022 Qitian Wu, Hengrui Zhang, Junchi Yan, David Wipf

There is increasing evidence suggesting neural networks' sensitivity to distribution shifts, so that research on out-of-distribution (OOD) generalization comes into the spotlight.

valid

GNNRank: Learning Global Rankings from Pairwise Comparisons via Directed Graph Neural Networks

1 code implementation1 Feb 2022 Yixuan He, Quan Gan, David Wipf, Gesine Reinert, Junchi Yan, Mihai Cucuringu

In this paper, we introduce neural networks into the ranking recovery problem by proposing the so-called GNNRank, a trainable GNN-based framework with digraph embedding.

Inductive Bias

The KFIoU Loss for Rotated Object Detection

3 code implementations29 Jan 2022 Xue Yang, Yue Zhou, Gefan Zhang, Jirui Yang, Wentao Wang, Junchi Yan, Xiaopeng Zhang, Qi Tian

This is in contrast to recent Gaussian modeling based rotation detectors e. g. GWD loss and KLD loss that involve a human-specified distribution distance metric which require additional hyperparameter tuning that vary across datasets and detectors.

Object object-detection +1

Appearance and Structure Aware Robust Deep Visual Graph Matching: Attack, Defense and Beyond

1 code implementation CVPR 2022 Qibing Ren, Qingquan Bao, Runzhong Wang, Junchi Yan

We first show that an adversarial attack on keypoint localities and the hidden graphs can cause significant accuracy drop to deep GM models.

Ranked #6 on Graph Matching on PASCAL VOC (matching accuracy metric)

Adversarial Attack Data Augmentation +2

Optimal LED Spectral Multiplexing for NIR2RGB Translation

1 code implementation CVPR 2022 Lei Liu, Yuze Chen, Junchi Yan, Yinqiang Zheng

The industry practice for night video surveillance is to use auxiliary near-infrared (NIR) LED diodes, usually centered at 850nm or 940nm, for scene illumination.

Translation

A General Framework for Evaluating Robustness of Combinatorial Optimization Solvers on Graphs

no code implementations28 Dec 2021 Han Lu, Zenan Li, Runzhong Wang, Qibing Ren, Junchi Yan, Xiaokang Yang

Solving combinatorial optimization (CO) on graphs is among the fundamental tasks for upper-stream applications in data mining, machine learning and operations research.

Adversarial Attack Combinatorial Optimization

AlphaRotate: A Rotation Detection Benchmark using TensorFlow

1 code implementation12 Nov 2021 Xue Yang, Yue Zhou, Junchi Yan

AlphaRotate is an open-source Tensorflow benchmark for performing scalable rotation detection on various datasets.

On Joint Learning for Solving Placement and Routing in Chip Design

2 code implementations NeurIPS 2021 Ruoyu Cheng, Junchi Yan

To achieve end-to-end placement learning, we first propose a joint learning method termed by DeepPlace for the placement of macros and standard cells, by the integration of reinforcement learning with a gradient based optimization scheme.

reinforcement-learning Reinforcement Learning (RL)

ZARTS: On Zero-order Optimization for Neural Architecture Search

no code implementations10 Oct 2021 Xiaoxing Wang, Wenxuan Guo, Junchi Yan, Jianlin Su, Xiaokang Yang

Also, we search on the search space of DARTS to compare with peer methods, and our discovered architecture achieves 97. 54% accuracy on CIFAR-10 and 75. 7% top-1 accuracy on ImageNet, which are state-of-the-art performance.

Neural Architecture Search

Towards Open-World Feature Extrapolation: An Inductive Graph Learning Approach

1 code implementation NeurIPS 2021 Qitian Wu, Chenxiao Yang, Junchi Yan

We target open-world feature extrapolation problem where the feature space of input data goes through expansion and a model trained on partially observed features needs to handle new features in test data without further retraining.

Graph Learning

DAAS: Differentiable Architecture and Augmentation Policy Search

no code implementations30 Sep 2021 Xiaoxing Wang, Xiangxiang Chu, Junchi Yan, Xiaokang Yang

Neural architecture search (NAS) has been an active direction of automatic machine learning (Auto-ML), aiming to explore efficient network structures.

Data Augmentation Neural Architecture Search

On Learning to Solve Cardinality Constrained Combinatorial Optimization in One-Shot: A Re-parameterization Approach via Gumbel-Sinkhorn-TopK

no code implementations29 Sep 2021 Runzhong Wang, Li Shen, Yiting Chen, Junchi Yan, Xiaokang Yang, DaCheng Tao

Cardinality constrained combinatorial optimization requires selecting an optimal subset of $k$ elements, and it will be appealing to design data-driven algorithms that perform TopK selection over a probability distribution predicted by a neural network.

Combinatorial Optimization One-Shot Learning +1

Zero-CL: Instance and Feature decorrelation for negative-free symmetric contrastive learning

no code implementations ICLR 2022 Shaofeng Zhang, Feng Zhu, Junchi Yan, Rui Zhao, Xiaokang Yang

The proposed two methods (FCL, ICL) can be combined synthetically, called Zero-CL, where ``Zero'' means negative samples are \textbf{zero} relevant, which allows Zero-CL to completely discard negative pairs i. e., with \textbf{zero} negative samples.

Contrastive Learning

ESCo: Towards Provably Effective and Scalable Contrastive Representation Learning

no code implementations29 Sep 2021 Hengrui Zhang, Qitian Wu, Shaofeng Zhang, Junchi Yan, David Wipf, Philip S. Yu

In this paper, we propose ESCo (Effective and Scalable Contrastive), a new contrastive framework which is essentially an instantiation of the Information Bottleneck principle under self-supervised learning settings.

Contrastive Learning Representation Learning +1

Learning mixture of neural temporal point processes for event sequence clustering

no code implementations29 Sep 2021 Yunhao Zhang, Junchi Yan, Zhenyu Ren, Jian Yin

To fill the gap, we propose Mixture of Neural Temporal Point Processes (NTPP-MIX), a general framework that can utilize many existing NTPPs for event sequence clustering.

Clustering Point Processes

On the Expressiveness, Predictability and Interpretability of Neural Temporal Point Processes

no code implementations29 Sep 2021 Liangliang Shi, Fangyu Ding, Junchi Yan, Yanjie Duan, Guangjian Tian

Despite the fast advance in neural temporal point processes (NTPP) which enjoys high model capacity, there are still some standing gaps to fill including model expressiveness, predictability, and interpretability, especially with the wide application of event sequence modeling.

Point Processes

Adversarial Robustness via Adaptive Label Smoothing

no code implementations29 Sep 2021 Qibing Ren, Liangliang Shi, Lanjun Wang, Junchi Yan

We first show both theoretically and empirically that strong smoothing in AT increases local smoothness of the loss surface which is beneficial for robustness but sacrifices the training loss which influences the accuracy of samples near the decision boundary.

Adversarial Robustness

Improving Generative Adversarial Networks via Adversarial Learning in Latent Space

no code implementations29 Sep 2021 Yang Li, Yichuan Mo, Liangliang Shi, Junchi Yan, Xiaolu Zhang, Jun Zhou

Although many efforts have been made in terms of backbone architecture design, loss function, and training techniques, few results have been obtained on how the sampling in latent space can affect the final performance, and existing works on latent space mainly focus on controllability.

Nonlinear ICA Using Volume-Preserving Transformations

no code implementations ICLR 2022 Xiaojiang Yang, Yi Wang, Jiacheng Sun, Xing Zhang, Shifeng Zhang, Zhenguo Li, Junchi Yan

Nonlinear ICA is a fundamental problem in machine learning, aiming to identify the underlying independent components (sources) from data which is assumed to be a nonlinear function (mixing function) of these sources.

Learning Dual Dynamic Representations on Time-Sliced User-Item Interaction Graphs for Sequential Recommendation

1 code implementation24 Sep 2021 Zeyuan Chen, Wei zhang, Junchi Yan, Gang Wang, Jianyong Wang

Sequential Recommendation aims to recommend items that a target user will interact with in the near future based on the historically interacted items.

Representation Learning Sequential Recommendation

RSDet++: Point-based Modulated Loss for More Accurate Rotated Object Detection

1 code implementation24 Sep 2021 Wen Qian, Xue Yang, Silong Peng, Junchi Yan, Xiujuan Zhang

We classify the discontinuity of loss in both five-param and eight-param rotated object detection methods as rotation sensitivity error (RSE) which will result in performance degeneration.

Object object-detection +1

Dynamic Modeling of Hand-Object Interactions via Tactile Sensing

no code implementations9 Sep 2021 Qiang Zhang, Yunzhu Li, Yiyue Luo, Wan Shou, Michael Foshey, Junchi Yan, Joshua B. Tenenbaum, Wojciech Matusik, Antonio Torralba

This work takes a step on dynamics modeling in hand-object interactions from dense tactile sensing, which opens the door for future applications in activity learning, human-computer interactions, and imitation learning for robotics.

Contrastive Learning Imitation Learning +1

SelfSAGCN: Self-Supervised Semantic Alignment for Graph Convolution Network

1 code implementation CVPR 2021 Xu Yang, Cheng Deng, Zhiyuan Dang, Kun Wei, Junchi Yan

Specifically, the Identity Aggregation is applied to extract semantic features from labeled nodes, the Semantic Alignment is utilized to align node features obtained from different aspects using the class central similarity.

Representation Learning

Learning High-Precision Bounding Box for Rotated Object Detection via Kullback-Leibler Divergence

2 code implementations NeurIPS 2021 Xue Yang, Xiaojiang Yang, Jirui Yang, Qi Ming, Wentao Wang, Qi Tian, Junchi Yan

Taking the perspective that horizontal detection is a special case for rotated object detection, in this paper, we are motivated to change the design of rotation regression loss from induction paradigm to deduction methodology, in terms of the relation between rotation and horizontal detection.

Ranked #14 on Object Detection In Aerial Images on DOTA (using extra training data)

object-detection Object Detection In Aerial Images +1

IID-GAN: an IID Sampling Perspective for Regularizing Mode Collapse

no code implementations1 Jun 2021 Yang Li, Liangliang Shi, Junchi Yan

Based on this observation, considering a necessary condition of IID generation that the inverse samples from target data should also be IID in the source distribution, we propose a new loss to encourage the closeness between inverse samples of real data and the Gaussian source in latent space to regularize the generation to be IID from the target distribution.

Scalable and Explainable 1-Bit Matrix Completion via Graph Signal Learning

1 code implementation AAAI 2021 Chao Chen, Dongsheng Li, Junchi Yan, Hanchi Huang, Xiaokang Yang

One-bit matrix completion is an important class of positiveunlabeled (PU) learning problems where the observations consist of only positive examples, eg, in top-N recommender systems.

Collaborative Ranking Matrix Completion +1

ENPAR:Enhancing Entity and Entity Pair Representations for Joint Entity Relation Extraction

1 code implementation EACL 2021 Yijun Wang, Changzhi Sun, Yuanbin Wu, Hao Zhou, Lei LI, Junchi Yan

Current state-of-the-art systems for joint entity relation extraction (Luan et al., 2019; Wad-den et al., 2019) usually adopt the multi-task learning framework.

coreference-resolution Entity Typing +5

Tuning IR-cut Filter for Illumination-aware Spectral Reconstruction from RGB

no code implementations CVPR 2021 Bo Sun, Junchi Yan, Xiao Zhou, Yinqiang Zheng

To reconstruct spectral signals from multi-channel observations, in particular trichromatic RGBs, has recently emerged as a promising alternative to traditional scanning-based spectral imager.

Spectral Reconstruction

Generalizing Face Forgery Detection with High-frequency Features

no code implementations CVPR 2021 Yuchen Luo, Yong Zhang, Junchi Yan, Wei Liu

The second is the residual-guided spatial attention module that guides the low-level RGB feature extractor to concentrate more on forgery traces from a new perspective.

Vocal Bursts Intensity Prediction

Learning Comprehensive Motion Representation for Action Recognition

no code implementations23 Mar 2021 Mingyu Wu, Boyuan Jiang, Donghao Luo, Junchi Yan, Yabiao Wang, Ying Tai, Chengjie Wang, Jilin Li, Feiyue Huang, Xiaokang Yang

For action recognition learning, 2D CNN-based methods are efficient but may yield redundant features due to applying the same 2D convolution kernel to each frame.

Action Recognition

Anomaly Detection of Time Series with Smoothness-Inducing Sequential Variational Auto-Encoder

no code implementations2 Feb 2021 Longyuan Li, Junchi Yan, Haiyang Wang, Yaohui Jin

Our model is based on Variational Auto-Encoder (VAE), and its backbone is fulfilled by a Recurrent Neural Network to capture latent temporal structures of time series for both generative model and inference model.

Anomaly Detection Density Estimation +2

Learning Interpretable Deep State Space Model for Probabilistic Time Series Forecasting

no code implementations31 Jan 2021 Longyuan Li, Junchi Yan, Xiaokang Yang, Yaohui Jin

We propose a deep state space model for probabilistic time series forecasting whereby the non-linear emission model and transition model are parameterized by networks and the dependency is modeled by recurrent neural nets.

Decision Making Management +2

Rethinking Rotated Object Detection with Gaussian Wasserstein Distance Loss

2 code implementations28 Jan 2021 Xue Yang, Junchi Yan, Qi Ming, Wentao Wang, Xiaopeng Zhang, Qi Tian

Boundary discontinuity and its inconsistency to the final detection metric have been the bottleneck for rotating detection regression loss design.

Ranked #16 on Object Detection In Aerial Images on DOTA (using extra training data)

object-detection Object Detection In Aerial Images +2

Cognitive Visual Inspection Service for LCD Manufacturing Industry

no code implementations11 Jan 2021 Yuanyuan Ding, Junchi Yan, Guoqiang Hu, Jun Zhu

This paper discloses a novel visual inspection system for liquid crystal display (LCD), which is currently a dominant type in the FPD industry.

Defect Detection

Self-supervised representation learning via adaptive hard-positive mining

no code implementations1 Jan 2021 Shaofeng Zhang, Junchi Yan, Xiaokang Yang

Despite their success in perception over the last decade, deep neural networks are also known ravenous to labeled data for training, which limits their applicability to real-world problems.

Contrastive Learning Representation Learning +1

InvertGAN: Reducing mode collapse with multi-dimensional Gaussian Inversion

no code implementations1 Jan 2021 Liangliang Shi, Yang Li, Junchi Yan

Generative adversarial networks have shown their ability in capturing high-dimensional complex distributions and generating realistic data samples e. g. images.

Self-supervised Disentangled Representation Learning

no code implementations1 Jan 2021 Xiaojiang Yang, Yitong Sun, Junchi Yan

In our experiments, we find that even the data is only augmented along a few latent variables, more latent variables can be identified, and adding a small noise in data space can stabilize this outcome.

Disentanglement Self-Supervised Learning

GSdyn: Learning training dynamics via online Gaussian optimization with gradient states

no code implementations1 Jan 2021 Haoran Liao, Junchi Yan, Zimin Feng

Bayesian optimization, whose efficiency for automatic hyperparameter tuning has been verified over the decade, still faces a standing dilemma between massive consumption of time and suboptimal search results.

Bayesian Optimization

Learning Latent Topology for Graph Matching

no code implementations1 Jan 2021 Tianshu Yu, Runzhong Wang, Junchi Yan, Baoxin Li

Graph matching (GM) has been traditionally modeled as a deterministic optimization problem characterized by an affinity matrix under pre-defined graph topology.

Graph Generation Graph Matching +1

Revocable Deep Reinforcement Learning with Affinity Regularization for Outlier-Robust Graph Matching

2 code implementations16 Dec 2020 Chang Liu, Zetian Jiang, Runzhong Wang, Junchi Yan, Lingxiao Huang, Pinyan Lu

As such, the agent can finish inlier matching timely when the affinity score stops growing, for which otherwise an additional parameter i. e. the number of inliers is needed to avoid matching outliers.

Combinatorial Optimization Decision Making +3

Adversarial Learning for Robust Deep Clustering

1 code implementation NeurIPS 2020 Xu Yang, Cheng Deng, Kun Wei, Junchi Yan, Wei Liu

Meanwhile, we devise an adversarial attack strategy to explore samples that easily fool the clustering layers but do not impact the performance of the deep embedding.

Adversarial Attack Clustering +1

The Diversified Ensemble Neural Network

no code implementations NeurIPS 2020 Shaofeng Zhang, Meng Liu, Junchi Yan

Ensemble is a general way of improving the accuracy and stability of learning models, especially for the generalization ability on small datasets.

Graduated Assignment for Joint Multi-Graph Matching and Clustering with Application to Unsupervised Graph Matching Network Learning

1 code implementation NeurIPS 2020 Runzhong Wang, Junchi Yan, Xiaokang Yang

This paper considers the setting of jointly matching and clustering multiple graphs belonging to different groups, which naturally rises in many realistic problems.

Clustering Graph Matching

Combinatorial Learning of Graph Edit Distance via Dynamic Embedding

no code implementations CVPR 2021 Runzhong Wang, Tianqi Zhang, Tianshu Yu, Junchi Yan, Xiaokang Yang

This paper presents a hybrid approach by combing the interpretability of traditional search-based techniques for producing the edit path, as well as the efficiency and adaptivity of deep embedding models to achieve a cost-effective GED solver.

ROME: Robustifying Memory-Efficient NAS via Topology Disentanglement and Gradient Accumulation

no code implementations ICCV 2023 Xiaoxing Wang, Xiangxiang Chu, Yuda Fan, Zhexi Zhang, Bo Zhang, Xiaokang Yang, Junchi Yan

Albeit being a prevalent architecture searching approach, differentiable architecture search (DARTS) is largely hindered by its substantial memory cost since the entire supernet resides in the memory.

Disentanglement Neural Architecture Search

Dense Label Encoding for Boundary Discontinuity Free Rotation Detection

3 code implementations CVPR 2021 Xue Yang, Liping Hou, Yue Zhou, Wentao Wang, Junchi Yan

Rotation detection serves as a fundamental building block in many visual applications involving aerial image, scene text, and face etc.

Ranked #29 on Object Detection In Aerial Images on DOTA (using extra training data)

Classification General Classification +2

Lightweight Single-Image Super-Resolution Network with Attentive Auxiliary Feature Learning

1 code implementation13 Nov 2020 Xuehui Wang, Qing Wang, Yuzhi Zhao, Junchi Yan, Lei Fan, Long Chen

In this paper, we develop a computation efficient yet accurate network based on the proposed attentive auxiliary features (A$^2$F) for SISR.

Image Super-Resolution

Learning Calibrated Uncertainties for Domain Shift: A Distributionally Robust Learning Approach

no code implementations8 Oct 2020 Haoxuan Wang, Zhiding Yu, Yisong Yue, Anima Anandkumar, Anqi Liu, Junchi Yan

We propose a framework for learning calibrated uncertainties under domain shifts, where the source (training) distribution differs from the target (test) distribution.

Density Ratio Estimation Unsupervised Domain Adaptation

Fusing Motion Patterns and Key Visual Information for Semantic Event Recognition in Basketball Videos

no code implementations13 Jul 2020 Lifang Wu, Zhou Yang, Qi. Wang, Meng Jian, Boxuan Zhao, Junchi Yan, Chang Wen Chen

Based on the observations, we propose a scheme to fuse global and local motion patterns (MPs) and key visual information (KVI) for semantic event recognition in basketball videos.

Group Activity Recognition Optical Flow Estimation

Towards Open-World Recommendation: An Inductive Model-based Collaborative Filtering Approach

1 code implementation9 Jul 2020 Qitian Wu, Hengrui Zhang, Xiaofeng Gao, Junchi Yan, Hongyuan Zha

The first model follows conventional matrix factorization which factorizes a group of key users' rating matrix to obtain meta latents.

Collaborative Filtering Matrix Completion +2

Permutation Matters: Anisotropic Convolutional Layer for Learning on Point Clouds

1 code implementation27 May 2020 Zhongpai Gao, Guangtao Zhai, Junchi Yan, Xiaokang Yang

Various point neural networks have been developed with isotropic filters or using weighting matrices to overcome the structure inconsistency on point clouds.

Representation Learning Semantic Segmentation

F2A2: Flexible Fully-decentralized Approximate Actor-critic for Cooperative Multi-agent Reinforcement Learning

no code implementations17 Apr 2020 Wenhao Li, Bo Jin, Xiangfeng Wang, Junchi Yan, Hongyuan Zha

Traditional centralized multi-agent reinforcement learning (MARL) algorithms are sometimes unpractical in complicated applications, due to non-interactivity between agents, curse of dimensionality and computation complexity.

Multi-agent Reinforcement Learning Reinforcement Learning (RL) +2

GFTE: Graph-based Financial Table Extraction

1 code implementation17 Mar 2020 Yiren Li, Zheng Huang, Junchi Yan, Yi Zhou, Fan Ye, Xianhui Liu

Tabular data is a crucial form of information expression, which can organize data in a standard structure for easy information retrieval and comparison.

Information Retrieval Position +2

On the Arbitrary-Oriented Object Detection: Classification based Approaches Revisited

4 code implementations ECCV 2020 Xue Yang, Junchi Yan

For the resulting circularly distributed angle classification problem, we first devise a Circular Smooth Label technique to handle the periodicity of angle and increase the error tolerance to adjacent angles.

Ranked #37 on Object Detection In Aerial Images on DOTA (using extra training data)

Classification General Classification +4

HMRL: Hyper-Meta Learning for Sparse Reward Reinforcement Learning Problem

no code implementations11 Feb 2020 Yun Hua, Xiangfeng Wang, Bo Jin, Wenhao Li, Junchi Yan, Xiaofeng He, Hongyuan Zha

In spite of the success of existing meta reinforcement learning methods, they still have difficulty in learning a meta policy effectively for RL problems with sparse reward.

Meta-Learning Meta Reinforcement Learning +2

Learning deep graph matching with channel-independent embedding and Hungarian attention

no code implementations ICLR 2020 Tianshu Yu, Runzhong Wang, Junchi Yan, Baoxin Li

Graph matching aims to establishing node-wise correspondence between two graphs, which is a classic combinatorial problem and in general NP-complete.

Ranked #15 on Graph Matching on PASCAL VOC (matching accuracy metric)

Graph Matching Hard Attention

Towards Better Understanding of Disentangled Representations via Mutual Information

no code implementations25 Nov 2019 Xiaojiang Yang, Wendong Bi, Yitong Sun, Yu Cheng, Junchi Yan

Most existing works on disentangled representation learning are solely built upon an marginal independence assumption: all factors in disentangled representations should be statistically independent.

Disentanglement Inductive Bias +1

Decoding Spiking Mechanism with Dynamic Learning on Neuron Population

no code implementations21 Nov 2019 Zhijie Chen, Junchi Yan, Longyuan Li, Xiaokang Yang

Our model is aimed to reconstruct neuron information while inferring representations of neuron spiking states.

Heterogeneous Graph-based Knowledge Transfer for Generalized Zero-shot Learning

no code implementations20 Nov 2019 Jun-Jie Wang, Xiangfeng Wang, Bo Jin, Junchi Yan, Wenjie Zhang, Hongyuan Zha

To this end, we propose a novel heterogeneous graph-based knowledge transfer method (HGKT) for GZSL, agnostic to unseen classes and instances, by leveraging graph neural network.

Generalized Zero-Shot Learning Transfer Learning

Learning Modulated Loss for Rotated Object Detection

2 code implementations19 Nov 2019 Wen Qian, Xue Yang, Silong Peng, Yue Guo, Junchi Yan

Popular rotated detection methods usually use five parameters (coordinates of the central point, width, height, and rotation angle) to describe the rotated bounding box and l1-loss as the loss function.

Ranked #43 on Object Detection In Aerial Images on DOTA (using extra training data)

Object object-detection +2

Pingan Smart Health and SJTU at COIN - Shared Task: utilizing Pre-trained Language Models and Common-sense Knowledge in Machine Reading Tasks

no code implementations WS 2019 Xiepeng Li, Zhexi Zhang, Wei Zhu, Zheng Li, Yuan Ni, Peng Gao, Junchi Yan, Guotong Xie

We have experimented both (a) improving the fine-tuning of pre-trained language models on a task with a small dataset size, by leveraging datasets of similar tasks; and (b) incorporating the distributional representations of a KG onto the representations of pre-trained language models, via simply concatenation or multi-head attention.

Common Sense Reasoning Machine Reading Comprehension +2

R3Det: Refined Single-Stage Detector with Feature Refinement for Rotating Object

10 code implementations15 Aug 2019 Xue Yang, Junchi Yan, Ziming Feng, Tao He

Considering the shortcoming of feature misalignment in existing refined single-stage detector, we design a feature refinement module to improve detection performance by getting more accurate features.

object-detection Object Detection In Aerial Images

Deep Spectral Clustering using Dual Autoencoder Network

no code implementations CVPR 2019 Xu Yang, Cheng Deng, Feng Zheng, Junchi Yan, Wei Liu

In this paper, we propose a joint learning framework for discriminative embedding and spectral clustering.

Clustering Deep Clustering +1

Learning Combinatorial Embedding Networks for Deep Graph Matching

1 code implementation ICCV 2019 Runzhong Wang, Junchi Yan, Xiaokang Yang

In addition with its NP-completeness nature, another important challenge is effective modeling of the node-wise and structure-wise affinity across graphs and the resulting objective, to guide the matching procedure effectively finding the true matching against noises.

Graph Embedding Graph Matching

Generalizing Graph Matching beyond Quadratic Assignment Model

no code implementations NeurIPS 2018 Tianshu Yu, Junchi Yan, Yilin Wang, Wei Liu, Baoxin Li

Graph matching has received persistent attention over decades, which can be formulated as a quadratic assignment problem (QAP).

Graph Matching

SCRDet: Towards More Robust Detection for Small, Cluttered and Rotated Objects

3 code implementations ICCV 2019 Xue Yang, Jirui Yang, Junchi Yan, Yue Zhang, Tengfei Zhang, Zhi Guo, Sun Xian, Kun fu

Specifically, a sampling fusion network is devised which fuses multi-layer feature with effective anchor sampling, to improve the sensitivity to small objects.

Ranked #47 on Object Detection In Aerial Images on DOTA (using extra training data)

object-detection Object Detection In Aerial Images

Collaborative Filtering with Stability

no code implementations6 Nov 2018 Dongsheng Li, Chao Chen, Qin Lv, Junchi Yan, Li Shang, Stephen M. Chu

Collaborative filtering (CF) is a popular technique in today's recommender systems, and matrix approximation-based CF methods have achieved great success in both rating prediction and top-N recommendation tasks.

Collaborative Filtering Recommendation Systems

Incremental Multi-graph Matching via Diversity and Randomness based Graph Clustering

no code implementations ECCV 2018 Tianshu Yu, Junchi Yan, Wei Liu, Baoxin Li

In this paper, we present an incremental multi-graph matching approach, which deals with the arriving graph utilizing the previous matching results under the global consistency constraint.

Clustering Graph Clustering +1

Decoupled Learning for Factorial Marked Temporal Point Processes

no code implementations21 Jan 2018 Weichang Wu, Junchi Yan, Xiaokang Yang, Hongyuan Zha

In conventional (multi-dimensional) marked temporal point process models, event is often encoded by a single discrete variable i. e. a marker.

Point Processes

tau-FPL: Tolerance-Constrained Learning in Linear Time

no code implementations15 Jan 2018 Ao Zhang, Nan Li, Jian Pu, Jun Wang, Junchi Yan, Hongyuan Zha

Learning a classifier with control on the false-positive rate plays a critical role in many machine learning applications.

Attention based convolutional neural network for predicting RNA-protein binding sites

no code implementations6 Dec 2017 Xiaoyong Pan, Junchi Yan

In this study, we present an attention based convolutional neural network, iDeepA, to predict RNA-protein binding sites from raw RNA sequences.

Joint Cuts and Matching of Partitions in One Graph

no code implementations CVPR 2018 Tianshu Yu, Junchi Yan, Jieyi Zhao, Baoxin Li

As two fundamental problems, graph cuts and graph matching have been investigated over decades, resulting in vast literature in these two topics respectively.

Graph Matching

Transductive Non-linear Learning for Chinese Hypernym Prediction

no code implementations ACL 2017 Chengyu Wang, Junchi Yan, Aoying Zhou, Xiaofeng He

Finding the correct hypernyms for entities is essential for taxonomy learning, fine-grained entity categorization, query understanding, etc.

Relation Extraction Transductive Learning

Modeling The Intensity Function Of Point Process Via Recurrent Neural Networks

2 code implementations24 May 2017 Shuai Xiao, Junchi Yan, Stephen M. Chu, Xiaokang Yang, Hongyuan Zha

In this paper, we model the background by a Recurrent Neural Network (RNN) with its units aligned with time series indexes while the history effect is modeled by another RNN whose units are aligned with asynchronous events to capture the long-range dynamics.

Point Processes Time Series +1

Wasserstein Learning of Deep Generative Point Process Models

1 code implementation NeurIPS 2017 Shuai Xiao, Mehrdad Farajtabar, Xiaojing Ye, Junchi Yan, Le Song, Hongyuan Zha

Point processes are becoming very popular in modeling asynchronous sequential data due to their sound mathematical foundation and strength in modeling a variety of real-world phenomena.

Point Processes

Deep Extreme Multi-label Learning

1 code implementation12 Apr 2017 Wenjie Zhang, Junchi Yan, Xiangfeng Wang, Hongyuan Zha

Extreme multi-label learning (XML) or classification has been a practical and important problem since the boom of big data.

Classification Extreme Multi-Label Classification +2

Learning Correspondence Structures for Person Re-identification

no code implementations20 Mar 2017 Weiyao Lin, Yang shen, Junchi Yan, Mingliang Xu, Jianxin Wu, Jingdong Wang, Ke Lu

We first introduce a boosting-based approach to learn a correspondence structure which indicates the patch-wise matching probabilities between images from a target camera pair.

Patch Matching Person Re-Identification

Negative-Unlabeled Tensor Factorization for Location Category Inference from Highly Inaccurate Mobility Data

no code implementations21 Feb 2017 Jinfeng Yi, Qi Lei, Wesley Gifford, Ji Liu, Junchi Yan

In order to efficiently solve the proposed framework, we propose a parameter-free and scalable optimization algorithm by effectively exploring the sparse and low-rank structure of the tensor.

A Tube-and-Droplet-based Approach for Representing and Analyzing Motion Trajectories

no code implementations10 Sep 2016 Weiyao Lin, Yang Zhou, Hongteng Xu, Junchi Yan, Mingliang Xu, Jianxin Wu, Zicheng Liu

Our approach first leverages the complete information from given trajectories to construct a thermal transfer field which provides a context-rich way to describe the global motion pattern in a scene.

3D Action Recognition Anomaly Detection +2

A constrained clustering based approach for matching a collection of feature sets

no code implementations12 Jun 2016 Junchi Yan, Zhe Ren, Hongyuan Zha, Stephen Chu

In this paper, we consider the problem of finding the feature correspondences among a collection of feature sets, by using their point-wise unary features.

Attribute Constrained Clustering +1

Ternary Weight Networks

5 code implementations16 May 2016 Fengfu Li, Bin Liu, Xiaoxing Wang, Bo Zhang, Junchi Yan

We present a memory and computation efficient ternary weight networks (TWNs) - with weights constrained to +1, 0 and -1.

Model Compression object-detection +1

Self-Paced Multi-Task Learning

no code implementations6 Apr 2016 Changsheng Li, Junchi Yan, Fan Wei, Weishan Dong, Qingshan Liu, Hongyuan Zha

In this paper, we propose a novel multi-task learning (MTL) framework, called Self-Paced Multi-Task Learning (SPMTL).

Multi-Task Learning

Cannot find the paper you are looking for? You can Submit a new open access paper.