no code implementations • 15 Apr 2024 • Hanjing Wang, Qiang Ji
Specifically, we propose a gradient-based approach to assess epistemic uncertainty, analyzing the gradients of outputs relative to model parameters, and thereby indicating necessary model adjustments to accurately represent the inputs.
no code implementations • 5 Apr 2024 • Yufei Zhang, Jeffrey O. Kephart, Zijun Cui, Qiang Ji
PhysPT exploits a Transformer encoder-decoder backbone to effectively learn human dynamics in a self-supervised manner.
no code implementations • 20 Dec 2023 • Naiyu Yin, Tian Gao, Yue Yu, Qiang Ji
We then propose an effective two-phase iterative DAG learning algorithm to address the increasing optimization difficulties and to learn a causal DAG from data with heteroscedastic variable noise under varying variance.
no code implementations • ICCV 2023 • Yufei Zhang, Hanjing Wang, Jeffrey O. Kephart, Qiang Ji
While 3D body reconstruction methods have made remarkable progress recently, it remains difficult to acquire the sufficiently accurate and numerous 3D supervisions required for training.
no code implementations • CVPR 2023 • Hanjing Wang, Dhiraj Joshi, Shiqiang Wang, Qiang Ji
Predictions made by deep learning models are prone to data perturbations, adversarial attacks, and out-of-distribution inputs.
no code implementations • ICCV 2023 • Hongji Guo, Qiang Ji
During the inference, the decoder is discared and a RNN-based classifier takes the output of the encoder for gait recognition.
no code implementations • CVPR 2023 • Zijun Cui, Chenyi Kuang, Tian Gao, Kartik Talamadupula, Qiang Ji
In this paper, we propose a biomechanics-guided AU detection approach, where facial muscle activation forces are modelled, and are employed to predict AU activation.
no code implementations • 30 Nov 2022 • Zijun Cui, Tian Gao, Kartik Talamadupula, Qiang Ji
Based on our taxonomy, we provide a systematic review of existing techniques, different from existing works that survey integration approaches agnostic to taxonomy of knowledge.
1 code implementation • CVPR 2023 • Bashirul Azam Biswas, Qiang Ji
The quality of scene graphs generated by the state-of-the-art (SOTA) models is compromised due to the long-tail nature of the relationships and their parent object pairs.
no code implementations • 16 Jun 2022 • Zijun Cui, Naiyu Yin, Yuru Wang, Qiang Ji
Causal discovery is to learn cause-effect relationships among variables given observational data and is important for many applications.
no code implementations • CVPR 2022 • Hongji Guo, Hanjing Wang, Qiang Ji
The model prediction uncertainty is used to improve both training and inference.
1 code implementation • 12 Aug 2021 • Shreya Ghosh, Abhinav Dhall, Munawar Hayat, Jarrod Knibbe, Qiang Ji
Eye gaze analysis is an important research problem in the field of Computer Vision and Human-Computer Interaction.
no code implementations • 25 Jun 2021 • Qiang Ji, Kang Wang
Model-based eye tracking, however, is susceptible to eye feature detection errors, in particular for eye tracking in the wild.
no code implementations • CVPR 2021 • Tengfei Song, Zijun Cui, Wenming Zheng, Qiang Ji
In this paper, we propose a novel hybrid message passing neural network with performance-driven structures (HMP-PS), which combines complementary message passing methods and captures more possible structures in a Bayesian manner.
no code implementations • CVPR 2021 • Tengfei Song, Zijun Cui, Yuru Wang, Wenming Zheng, Qiang Ji
Second, we introduce probabilistic graph convolution that allows to perform graph convolution on the distribution of Bayesian Network structure to extract AU structural features.
1 code implementation • 14 Jun 2021 • Yue Yu, Tian Gao, Naiyu Yin, Qiang Ji
To further improve efficiency, we propose a novel learning framework to model and learn the weighted adjacency matrices in the DAG space directly.
no code implementations • NeurIPS 2020 • Zijun Cui, Tengfei Song, Yuru Wang, Qiang Ji
This paper proposes to systematically capture their dependencies and incorporate them into a deep learning framework for joint facial expression recognition and action unit detection.
1 code implementation • NeurIPS 2019 • Lisha Chen, Hui Su, Qiang Ji
Existing deep learning based facial landmark detection methods have achieved excellent performance.
Ranked #6 on Facial Landmark Detection on 300W
no code implementations • 16 Sep 2020 • Zijun Cui, Pavan Kapanipathi, Kartik Talamadupula, Tian Gao, Qiang Ji
Knowledge graph completion (also known as relation prediction) is the task of inferring missing facts given existing ones.
no code implementations • 9 Mar 2020 • Claudiu Albulescu, Aviral Tiwari, Qiang Ji
In all pairs of commodity indexes, we find increased co-movements in extreme situations, a stronger dependence between energy and other commodity markets at lower tails, and a 'V-type' local dependence for the energy-metal pairs.
no code implementations • 3 Oct 2019 • Sicheng Zhao, Shangfei Wang, Mohammad Soleymani, Dhiraj Joshi, Qiang Ji
Affective computing (AC) of these data can help to understand human behaviors and enable wide applications.
no code implementations • 12 Mar 2019 • Chao Gou, Tianyu Shen, Wenbo Zheng, Huadan Xue, Hui Yu, Qiang Ji, Zhengyu Jin, Fei-Yue Wang
Artificial imaging systems are introduced to select and prescriptively generate medical image data in a knowledge-driven way to utilize medical domain knowledge.
no code implementations • CVPR 2018 • Yong Zhang, Rui Zhao, Wei-Ming Dong, Bao-Gang Hu, Qiang Ji
The majority of methods directly apply supervised learning techniques to AU intensity estimation while few methods exploit unlabeled samples to improve the performance.
no code implementations • CVPR 2018 • Yong Zhang, Wei-Ming Dong, Bao-Gang Hu, Qiang Ji
To alleviate this issue, we propose a knowledge-driven method for jointly learning multiple AU classifiers without any AU annotation by leveraging prior probabilities on AUs, including expression-independent and expression-dependent AU probabilities.
no code implementations • CVPR 2018 • Yong Zhang, Wei-Ming Dong, Bao-Gang Hu, Qiang Ji
Facial action unit (AU) intensity estimation plays an important role in affective computing and human-computer interaction.
no code implementations • CVPR 2018 • Kang Wang, Rui Zhao, Qiang Ji
Through a top-down inference, the HGM can synthesize eye images consistent with the given eye gaze.
no code implementations • 15 May 2018 • Yue Wu, Qiang Ji
The regression-based methods implicitly capture facial shape and appearance information.
no code implementations • 21 Oct 2017 • Siqi Nie, Ziheng Wang, Qiang Ji
A learning method is then proposed to perform efficient learning for the proposed model.
no code implementations • 13 Oct 2017 • Siqi Nie, Meng Zheng, Qiang Ji
The major difficulty of learning and inference with deep directed models with many latent variables is the intractable inference due to the dependencies among the latent variables and the exponential number of latent variable configurations.
no code implementations • ICCV 2017 • Kang Wang, Qiang Ji
The key idea is to leverage on the proposed 3D eye-face model, from which we can estimate 3D eye gaze from observed 2D facial landmarks.
no code implementations • ICCV 2017 • Quan Gan, Shangfei Wang, Longfei Hao, Qiang Ji
After that, a joint representation is extracted from the top layers of the two deep networks, and thus captures the high order dependencies between visual modality and audio modality.
no code implementations • ICCV 2017 • Shan Wu, Shangfei Wang, Bowen Pan, Qiang Ji
To address this, we propose a deep facial action unit recognition approach learning from partially AU-labeled data.
no code implementations • ICCV 2015 • Yue Wu, Qiang Ji
In this work, we propose a unified robust cascade regression framework that can handle both images with severe occlusion and images with large head poses.
no code implementations • CVPR 2016 • Yue Wu, Qiang Ji
Furthermore, we propose to exploit the target domain knowledge and incorporate such prior knowledge as a constraint during transfer learning to ensure that the transferred data satisfies certain properties of the target domain.
Facial Expression Recognition Facial Expression Recognition (FER) +1
no code implementations • CVPR 2017 • Yue Wu, Chao Gou, Qiang Ji
Facial landmark detection, head pose estimation, and facial deformation analysis are typical facial behavior analysis tasks in computer vision.
no code implementations • CVPR 2016 • Yue Wu, Qiang Ji
Experimental results demonstrate that the intertwined relationships of facial action units and face shapes boost the performances of both facial action unit recognition and facial landmark detection.
no code implementations • CVPR 2013 • Yue Wu, Zuoguan Wang, Qiang Ji
To handle pose variations, the frontal face shape prior model is incorporated into a 3-way RBM model that could capture the relationship between frontal face shapes and non-frontal face shapes.
no code implementations • CVPR 2014 • Yue Wu, Ziheng Wang, Qiang Ji
Facial feature detection from facial images has attracted great attention in the field of computer vision.
no code implementations • CVPR 2016 • Rui Zhao, Quan Gan, Shangfei Wang, Qiang Ji
In fully supervised case, all the frames are provided with intensity annotations.
no code implementations • CVPR 2016 • Zheng Zhang, Jeff M. Girard, Yue Wu, Xing Zhang, Peng Liu, Umur Ciftci, Shaun Canavan, Michael Reale, Andy Horowitz, Huiyuan Yang, Jeffrey F. Cohn, Qiang Ji, Lijun Yin
The corpus further includes derived features from 3D, 2D, and IR (infrared) sensors and baseline results for facial expression and action unit detection.
no code implementations • ICCV 2015 • Tian Gao, Ziheng Wang, Qiang Ji
Then we apply structured feature selection to two applications: 1) We introduce a new method that enables STMB to scale up and show the competitive performance of our algorithms on large-scale image classification tasks.
no code implementations • NeurIPS 2015 • Tian Gao, Qiang Ji
We focus on the discovery and identification of direct causes and effects of a target variable in a causal network.
no code implementations • 15 Jun 2015 • Siqi Nie, Qiang Ji
Deep directed generative models have attracted much attention recently due to their expressive representation power and the ability of ancestral sampling.
no code implementations • CVPR 2015 • Ziheng Wang, Qiang Ji
Experimental results on different applications demonstrate the effectiveness of the proposed methods for exploiting hidden information and their superior performance to existing methods.
no code implementations • CVPR 2015 • Xiaoyang Wang, Qiang Ji
Video event recognition still faces great challenges due to large intra-class variation and low image resolution, in particular for surveillance videos.
no code implementations • The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2014 2014 • Xiaoyang Wang, Qiang Ji
These three levels of context provide crucial bottom-up, middle level, and top down information that can benefit the recognition task itself.
Ranked #1 on Action Recognition on VIRAT Ground 2.0
no code implementations • NeurIPS 2014 • Siqi Nie, Denis Deratani Maua, Cassio Polpo de Campos, Qiang Ji
This work presents novel algorithms for learning Bayesian network structures with bounded treewidth.
no code implementations • CVPR 2014 • Eran Swears, Anthony Hoogs, Qiang Ji, Kim Boyer
Modeling interactions of multiple co-occurring objects in a complex activity is becoming increasingly popular in the video domain.
no code implementations • CVPR 2013 • Baoyuan Wu, Yifan Zhang, Bao-Gang Hu, Qiang Ji
As a result, many pairwise constraints between faces can be easily obtained from the temporal and spatial knowledge of the face tracks.
no code implementations • CVPR 2013 • Ziheng Wang, Shangfei Wang, Qiang Ji
Spatial-temporal relations among facial muscles carry crucial information about facial expressions yet have not been thoroughly exploited.
Facial Expression Recognition Facial Expression Recognition (FER)
no code implementations • NeurIPS 2012 • Zuoguan Wang, Siwei Lyu, Gerwin Schalk, Qiang Ji
In this work, we describe a new learning scheme for parametric learning, in which the target variables $\y$ can be modeled with a prior model $p(\y)$ and the relations between data and target variables are estimated through $p(\y)$ and a set of uncorresponded data $\x$ in training.