Qiang Ji
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Qiang Ji has a total of 148 co-authors. Here are the top co-authors by number of shared publications:
- Shangfei Wang
University of Science and Technology of China, Hefei, China
16 shared publications - Hanjing Wang
Rensselaer Polytechnic Institute
6 shared publications - Wanru Xu
Visual Intelligence + X International Cooperation Joint Laboratory of the Ministry of Education, Beijing, China
6 shared publications - Jeffrey O. Kephart
IBM Research
4 shared publications - Lili Wan
Institute of Information Science, Beijing Jiaotong University, Beijing, China
4 shared publications Shangfei Wang
4 shared publications
Latest Publications
- 9 September 2026
- Published by Springer Nature
- p. 224-242
Bayesian Uncertainty Attribution-Guided Fine-Tuning for Open-Set Action Recognition
- 9 September 2026
- Published by Springer Nature
- p. 224-242
Bayesian Uncertainty Attribution-Guided Fine-Tuning for Open-Set Action Recognition
- Open Access
Diffusion-based 3D Hand Motion Recovery with Intuitive Physics
While 3D hand reconstruction from monocular images has made significant progress, generating accurate and temporally coherent motion estimates from videos remains challenging, particularly during hand-object interactions. In this paper, we present a novel 3D hand motion recovery framework that enhances image-based reconstructions through a diffusion-based and physics-augmented motion refinement model. Our model captures the distribution of refined motion estimates conditioned on initial ones, generating improved sequences through an iterative denoising process. Instead of relying on scarce annotated video data, we train our model only using motion capture data without images. We identify valuable intuitive physics knowledge during hand-object interactions, including key motion states and their associated motion constraints. We effectively integrate these physical insights into our diffusion model to improve its performance. Extensive experiments demonstrate that our approach significantly improves various frame-wise reconstruction methods, achieving state-of-the-art (SOTA) performance on existing benchmarks.
Open AccessDiffusion-based 3D Hand Motion Recovery with Intuitive Physics
While 3D hand reconstruction from monocular images has made significant progress, generating accurate and temporally coherent motion estimates from videos remains challenging, particularly during hand-object interactions. In this paper, we present a novel 3D hand motion recovery framework that enhances image-based reconstructions through a diffusion-based and physics-augmented motion refinement model. Our model captures the distribution of refined motion estimates conditioned on initial ones, generating improved sequences through an iterative denoising process. Instead of relying on scarce annotated video data, we train our model only using motion capture data without images. We identify valuable intuitive physics knowledge during hand-object interactions, including key motion states and their associated motion constraints. We effectively integrate these physical insights into our diffusion model to improve its performance. Extensive experiments demonstrate that our approach significantly improves various frame-wise reconstruction methods, achieving state-of-the-art (SOTA) performance on existing benchmarks.
- 3 December 2024
- Published by Springer Nature
- p. 445-460
Efficient Nonlinear DAG Learning Under Projection Framework
- 3 December 2024
- Published by Springer Nature
- p. 445-460
Efficient Nonlinear DAG Learning Under Projection Framework
- 3 November 2024
- Published by Springer Nature
- p. 271-288
Integrating Markov Blanket Discovery Into Causal Representation Learning for Domain Generalization
- 3 November 2024
- Published by Springer Nature
- p. 271-288
Integrating Markov Blanket Discovery Into Causal Representation Learning for Domain Generalization
- 29 October 2024
- Published by Springer Nature
- p. 283-301
Bayesian Evidential Deep Learning for Online Action Detection
- 29 October 2024
- Published by Springer Nature
- p. 283-301
Bayesian Evidential Deep Learning for Online Action Detection
- 25 October 2024
- Published by Springer Nature
- p. 106-125
Weakly-Supervised 3D Hand Reconstruction with Knowledge Prior and Uncertainty Guidance
- 25 October 2024
- Published by Springer Nature
- p. 106-125
Weakly-Supervised 3D Hand Reconstruction with Knowledge Prior and Uncertainty Guidance
- 25 October 2024
- Published by Springer Nature
- p. 237-253
Optimization-Based Uncertainty Attribution Via Learning Informative Perturbations
- 25 October 2024
- Published by Springer Nature
- p. 237-253
Optimization-Based Uncertainty Attribution Via Learning Informative Perturbations
- Open Access
- 11 July 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Multimedia Computing, Communications, and Applications
- Vol. 22 (9) , 1-22
Boosting Semi-Supervised Video Captioning via Learning Candidates Adjusters
Video captioning is a multimodal task on both CV and NLP, whose goal is to automatically obtain the description of video content with natural language statements. Although there are amounts of video data, their annotations with description sentences are very limited. In this paper, we define the semi-supervised video captioning (SSVC) problem in order to improve performance with limited annotations by leveraging the semantic knowledge from both well-annotated samples and no-annotated samples. To address the problem, we introduce a LCA-boosted model (LCABM) for boosting SSVC, where it is first to explore a learnable candidates adjuster to adjust the caption candidates and then treat these adjusted captions as pesudo labels to train the SSVC model with no-annotated samples in reverse. In particular, the model learning is considered as a bi-level optimization problem and solved by an EM-like multi-stage training algorithm. The experiments show the effectiveness of our proposed LCABM, whose performance is comparable and even better than those state-of-the-art fully-supervised methods even with less annotations.
- 11 July 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Multimedia Computing, Communications, and Applications
- Vol. 22 (9) , 1-22
Open AccessBoosting Semi-Supervised Video Captioning via Learning Candidates Adjusters
Video captioning is a multimodal task on both CV and NLP, whose goal is to automatically obtain the description of video content with natural language statements. Although there are amounts of video data, their annotations with description sentences are very limited. In this paper, we define the semi-supervised video captioning (SSVC) problem in order to improve performance with limited annotations by leveraging the semantic knowledge from both well-annotated samples and no-annotated samples. To address the problem, we introduce a LCA-boosted model (LCABM) for boosting SSVC, where it is first to explore a learnable candidates adjuster to adjust the caption candidates and then treat these adjusted captions as pesudo labels to train the SSVC model with no-annotated samples in reverse. In particular, the model learning is considered as a bi-level optimization problem and solved by an EM-like multi-stage training algorithm. The experiments show the effectiveness of our proposed LCABM, whose performance is comparable and even better than those state-of-the-art fully-supervised methods even with less annotations.
- 24 April 2024
- Published by Springer Nature
- p. 101-112
Semantic Attribution for Explainable Uncertainty Quantification
- 24 April 2024
- Published by Springer Nature
- p. 101-112
Semantic Attribution for Explainable Uncertainty Quantification
- Open Access
Epistemic Uncertainty Quantification For Pre-trained Neural Network
Epistemic uncertainty quantification (UQ) identifies where models lack knowledge. Traditional UQ methods, often based on Bayesian neural networks, are not suitable for pre-trained non-Bayesian models. Our study addresses quantifying epistemic uncertainty for any pre-trained model, which does not need the original training data or model modifications and can ensure broad applicability regardless of network architectures or training techniques. 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. We first explore theoretical guarantees of gradient-based methods for epistemic UQ, questioning the view that this uncertainty is only calculable through differences between multiple models. We further improve gradient-driven UQ by using class-specific weights for integrating gradients and emphasizing distinct contributions from neural network layers. Additionally, we enhance UQ accuracy by combining gradient and perturbation methods to refine the gradients. We evaluate our approach on out-of-distribution detection, uncertainty calibration, and active learning, demonstrating its superiority over current state-of-the-art UQ methods for pre-trained models.
Open AccessEpistemic Uncertainty Quantification For Pre-trained Neural Network
Epistemic uncertainty quantification (UQ) identifies where models lack knowledge. Traditional UQ methods, often based on Bayesian neural networks, are not suitable for pre-trained non-Bayesian models. Our study addresses quantifying epistemic uncertainty for any pre-trained model, which does not need the original training data or model modifications and can ensure broad applicability regardless of network architectures or training techniques. 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. We first explore theoretical guarantees of gradient-based methods for epistemic UQ, questioning the view that this uncertainty is only calculable through differences between multiple models. We further improve gradient-driven UQ by using class-specific weights for integrating gradients and emphasizing distinct contributions from neural network layers. Additionally, we enhance UQ accuracy by combining gradient and perturbation methods to refine the gradients. We evaluate our approach on out-of-distribution detection, uncertainty calibration, and active learning, demonstrating its superiority over current state-of-the-art UQ methods for pre-trained models.