Search Results for author:
Found 42 papers, 25 papers with code
Orchestrate Latent Expertise: Advancing Online Continual Learning with Multi-Level Supervision and Reverse Self-Distillation
However, a notable gap from CL to OCL stems from the additional overfitting-underfitting dilemma associated with the use of rehearsal buffers: the inadequate learning of new training samples (underfitting) and the repeated learning of a few old training samples (overfitting).
ShapeLLM: Universal 3D Object Understanding for Embodied Interaction
This paper presents ShapeLLM, the first 3D Multimodal Large Language Model (LLM) designed for embodied interaction, exploring a universal 3D object understanding with 3D point clouds and languages.
Ranked #1 on
3D Object Captioning
on Objaverse
3D Object Captioning
3D Point Cloud Linear Classification
+10
DreamLLM: Synergistic Multimodal Comprehension and Creation
This paper presents DreamLLM, a learning framework that first achieves versatile Multimodal Large Language Models (MLLMs) empowered with frequently overlooked synergy between multimodal comprehension and creation.
Ranked #1 on
Visual Question Answering
on MMBench
(GPT-3.5 score metric)
VPP: Efficient Conditional 3D Generation via Voxel-Point Progressive Representation
VPP leverages structured voxel representation in the proposed Voxel Semantic Generator and the sparsity of unstructured point representation in the Point Upsampler, enabling efficient generation of multi-category objects.
Point-GCC: Universal Self-supervised 3D Scene Pre-training via Geometry-Color Contrast
Geometry and color information provided by the point clouds are both crucial for 3D scene understanding.
Ranked #1 on
Unsupervised 3D Semantic Segmentation
on ScanNetV2
Revisiting Data Augmentation in Model Compression: An Empirical and Comprehensive Study
In this paper, we revisit the usage of data augmentation in model compression and give a comprehensive study on the relation between model sizes and their optimal data augmentation policy.
CORSD: Class-Oriented Relational Self Distillation
Knowledge distillation conducts an effective model compression method while holding some limitations:(1) the feature based distillation methods only focus on distilling the feature map but are lack of transferring the relation of data examples; (2) the relational distillation methods are either limited to the handcrafted functions for relation extraction, such as L2 norm, or weak in inter- and intra- class relation modeling.
CLIP-FO3D: Learning Free Open-world 3D Scene Representations from 2D Dense CLIP
Training a 3D scene understanding model requires complicated human annotations, which are laborious to collect and result in a model only encoding close-set object semantics.
Hebbian and Gradient-based Plasticity Enables Robust Memory and Rapid Learning in RNNs
Comparing different plasticity rules under the same framework shows that Hebbian plasticity is well-suited for several memory and associative learning tasks; however, it is outperformed by gradient-based plasticity on few-shot regression tasks which require the model to infer the underlying mapping.
Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretraining
This motivates us to learn 3D representations by sharing the merits of both paradigms, which is non-trivial due to the pattern difference between the two paradigms.
Ranked #1 on
Zero-Shot Transfer 3D Point Cloud Classification
on ModelNet10
(using extra training data)
3D Point Cloud Linear Classification
Few-Shot 3D Point Cloud Classification
+2
Tiny Updater: Towards Efficient Neural Network-Driven Software Updating
Significant advancements have been accomplished with deep neural networks in diverse visual tasks, which have substantially elevated their deployment in edge device software.
Autoencoders as Cross-Modal Teachers: Can Pretrained 2D Image Transformers Help 3D Representation Learning?
The success of deep learning heavily relies on large-scale data with comprehensive labels, which is more expensive and time-consuming to fetch in 3D compared to 2D images or natural languages.
Ranked #5 on
Few-Shot 3D Point Cloud Classification
on ModelNet40 10-way (10-shot)
(using extra training data)
Few-Shot 3D Point Cloud Classification
Knowledge Distillation
+1
Language-Assisted 3D Feature Learning for Semantic Scene Understanding
To guide 3D feature learning toward important geometric attributes and scene context, we explore the help of textual scene descriptions.
Structured Knowledge Distillation Towards Efficient and Compact Multi-View 3D Detection
Detecting 3D objects from multi-view images is a fundamental problem in 3D computer vision.
LW-ISP: A Lightweight Model with ISP and Deep Learning
The deep learning (DL)-based methods of low-level tasks have many advantages over the traditional camera in terms of hardware prospects, error accumulation and imaging effects.
Contrastive Deep Supervision
The success of deep learning is usually accompanied by the growth in neural network depth.
Rethinking the Augmentation Module in Contrastive Learning: Learning Hierarchical Augmentation Invariance with Expanded Views
A data augmentation module is utilized in contrastive learning to transform the given data example into two views, which is considered essential and irreplaceable.
Region-aware Knowledge Distillation for Efficient Image-to-Image Translation
In this paper, we propose Region-aware Knowledge Distillation ReKo to compress image-to-image translation models.
PointDistiller: Structured Knowledge Distillation Towards Efficient and Compact 3D Detection
The remarkable breakthroughs in point cloud representation learning have boosted their usage in real-world applications such as self-driving cars and virtual reality.
Learn from Unpaired Data for Image Restoration: A Variational Bayes Approach
Current approaches aim at generating synthesized training data from unpaired samples by exploring the relationship between the corrupted and clean data.
Unsupervised Deep Learning Meets Chan-Vese Model
The Chan-Vese (CV) model is a classic region-based method in image segmentation.
Wavelet Knowledge Distillation: Towards Efficient Image-to-Image Translation
Instead of directly distilling the generated images of teachers, wavelet knowledge distillation first decomposes the images into different frequency bands with discrete wavelet transformation and then only distills the high frequency bands.
Finding the Task-Optimal Low-Bit Sub-Distribution in Deep Neural Networks
Besides, an efficient deployment flow for the mobile CPU is developed, achieving up to 7. 46$\times$ inference acceleration on an octa-core ARM CPU.
Multi-Glimpse Network: A Robust and Efficient Classification Architecture based on Recurrent Downsampled Attention
Inspired by this observation, we propose an end-to-end trainable Multi-Glimpse Network (MGNet) which aims to tackle the challenges of high computation and the lack of robustness based on recurrent downsampled attention mechanism.
AFEC: Active Forgetting of Negative Transfer in Continual Learning
Without accessing to the old training samples, knowledge transfer from the old tasks to each new task is difficult to determine, which might be either positive or negative.
Learning From Unpaired Data: A Variational Bayes Approach
Collecting the paired training data is a difficult task in practice, but the unpaired samples broadly exist.
Not All Regions are Worthy to be Distilled: Region-aware Knowledge Distillation Towards Efficient Image-to-Image Translation
To tackle this challenge, in this paper, we propose Region-aware Knowledge Distillation which first localizes the crucial regions in the images with attention mechanism.
An Unsupervised Deep Learning Approach for Real-World Image Denoising
In the real-world case, the noise distribution is so complex that the simplified additive white Gaussian (AWGN) assumption rarely holds, which significantly deteriorates the Gaussian denoisers' performance.
Improve Object Detection with Feature-based Knowledge Distillation: Towards Accurate and Efficient Detectors
In this paper, we suggest that the failure of knowledge distillation on object detection is mainly caused by two reasons: (1) the imbalance between pixels of foreground and background and (2) lack of distillation on the relation between different pixels.
Task-Oriented Feature Distillation
Moreover, an orthogonal loss is applied to the feature resizing layer in TOFD to improve the performance of knowledge distillation.
PCNN: Pattern-based Fine-Grained Regular Pruning towards Optimizing CNN Accelerators
Weight pruning is a powerful technique to realize model compression.
An Image Enhancing Pattern-based Sparsity for Real-time Inference on Mobile Devices
Weight pruning has been widely acknowledged as a straightforward and effective method to eliminate redundancy in Deep Neural Networks (DNN), thereby achieving acceleration on various platforms.
Light-weight Calibrator: a Separable Component for Unsupervised Domain Adaptation
Existing domain adaptation methods aim at learning features that can be generalized among domains.
Ranked #3 on
Domain Adaptation
on USPS-to-MNIST
Exploring Frequency Domain Interpretation of Convolutional Neural Networks
To the best of our knowledge, there is no study on the interpretation of modern CNNs from the perspective of the frequency proportion of filters.
PCONV: The Missing but Desirable Sparsity in DNN Weight Pruning for Real-time Execution on Mobile Devices
Model compression techniques on Deep Neural Network (DNN) have been widely acknowledged as an effective way to achieve acceleration on a variety of platforms, and DNN weight pruning is a straightforward and effective method.
Non-Structured DNN Weight Pruning -- Is It Beneficial in Any Platform?
Based on the proposed comparison framework, with the same accuracy and quantization, the results show that non-structrued pruning is not competitive in terms of both storage and computation efficiency.
Brain-inspired reverse adversarial examples
On contrast, current state-of-the-art deep learning approaches heavily depend on the variety of training samples and the capacity of the network.
SCAN: A Scalable Neural Networks Framework Towards Compact and Efficient Models
Remarkable achievements have been attained by deep neural networks in various applications.
Be Your Own Teacher: Improve the Performance of Convolutional Neural Networks via Self Distillation
Different from traditional knowledge distillation - a knowledge transformation methodology among networks, which forces student neural networks to approximate the softmax layer outputs of pre-trained teacher neural networks, the proposed self distillation framework distills knowledge within network itself.
Toward Extremely Low Bit and Lossless Accuracy in DNNs with Progressive ADMM
Weight quantization is one of the most important techniques of Deep Neural Networks (DNNs) model compression method.
Adversarial Robustness vs Model Compression, or Both?
Furthermore, this work studies two hypotheses about weight pruning in the conventional setting and finds that weight pruning is essential for reducing the network model size in the adversarial setting, training a small model from scratch even with inherited initialization from the large model cannot achieve both adversarial robustness and high standard accuracy.
StructADMM: A Systematic, High-Efficiency Framework of Structured Weight Pruning for DNNs
Without loss of accuracy on the AlexNet model, we achieve 2. 58X and 3. 65X average measured speedup on two GPUs, clearly outperforming the prior work.
