Deng Cai

papers

1,425

total citations

papers (27)

Learning to Break the Loop: Analyzing and Mitigating Repetitions for Neural Text Generation

NEURIPS 2022arXiv

citations

Graph R-CNN: Towards Accurate 3D Object Detection with Semantic-Decorated Local Graph

ECCV 2022arXiv

citations

DID-M3D: Decoupling Instance Depth for Monocular 3D Object Detection

ECCV 2022arXiv

citations

Learning To Affiliate: Mutual Centralized Learning for Few-Shot Classification

CVPR 2022arXiv

citations

Lidar Point Cloud Guided Monocular 3D Object Detection

ECCV 2022arXiv

citations

MonoNeRD: NeRF-like Representations for Monocular 3D Object Detection

ICCV 2023arXiv

citations

One-shot Implicit Animatable Avatars with Model-based Priors

ICCV 2023arXiv

citations

Towards Efficient Adversarial Training on Vision Transformers

ECCV 2022arXiv

citations

Frame-Wise Action Representations for Long Videos via Sequence Contrastive Learning

CVPR 2022arXiv

citations

Learning Occupancy for Monocular 3D Object Detection

CVPR 2024arXiv

citations

Repetition In Repetition Out: Towards Understanding Neural Text Degeneration from the Data Perspective

NEURIPS 2023arXiv

citations

TagCLIP: A Local-to-Global Framework to Enhance Open-Vocabulary Multi-Label Classification of CLIP without Training

AAAI 2024arXiv

citations

Salient Object Ranking With Position-Preserved Attention

ICCV 2021arXiv

citations

PersonalVideo: High ID-Fidelity Video Customization without Dynamic and Semantic Degradation

ICCV 2025arXiv

citations

3426 Regulating Intermediate 3D Features for Vision-Centric Autonomous Driving

AAAI 2024

citations

TASeg: Temporal Aggregation Network for LiDAR Semantic Segmentation

CVPR 2024arXiv

citations

MagicID: Hybrid Preference Optimization for ID-Consistent and Dynamic-Preserved Video Customization

ICCV 2025arXiv

citations

Retrieval is Accurate Generation

ICLR 2024arXiv

citations

Measuring and Reducing Model Update Regression in Structured Prediction for NLP

NEURIPS 2022arXiv

citations

Semi-supervised 3D Object Detection with PatchTeacher and PillarMix

AAAI 2024arXiv

citations

Pseudo Label Refinery for Unsupervised Domain Adaptation on Cross-dataset 3D Object Detection

CVPR 2024arXiv

citations

Towards Fine-Grained HBOE with Rendered Orientation Set and Laplace Smoothing

AAAI 2024

citations

STraj: Self-training for Bridging the Cross-Geography Gap in Trajectory Prediction

AAAI 2025

citations

Empowering Self-Learning of LLMs: Inner Knowledge Explicitation as a Catalyst

AAAI 2025

citations

Deng Cai

papers (27)

CLRNet: Cross Layer Refinement Network for Lane Detection

Sparse Fuse Dense: Towards High Quality 3D Detection With Depth Completion

Do Wider Neural Networks Really Help Adversarial Robustness?

Learning to Break the Loop: Analyzing and Mitigating Repetitions for Neural Text Generation

Graph R-CNN: Towards Accurate 3D Object Detection with Semantic-Decorated Local Graph

DID-M3D: Decoupling Instance Depth for Monocular 3D Object Detection

Learning To Affiliate: Mutual Centralized Learning for Few-Shot Classification

Lidar Point Cloud Guided Monocular 3D Object Detection

MonoNeRD: NeRF-like Representations for Monocular 3D Object Detection

One-shot Implicit Animatable Avatars with Model-based Priors

Towards Efficient Adversarial Training on Vision Transformers

Frame-Wise Action Representations for Long Videos via Sequence Contrastive Learning

Learning Occupancy for Monocular 3D Object Detection

Repetition In Repetition Out: Towards Understanding Neural Text Degeneration from the Data Perspective

TagCLIP: A Local-to-Global Framework to Enhance Open-Vocabulary Multi-Label Classification of CLIP without Training

Salient Object Ranking With Position-Preserved Attention

PersonalVideo: High ID-Fidelity Video Customization without Dynamic and Semantic Degradation

3426 Regulating Intermediate 3D Features for Vision-Centric Autonomous Driving

TASeg: Temporal Aggregation Network for LiDAR Semantic Segmentation

MagicID: Hybrid Preference Optimization for ID-Consistent and Dynamic-Preserved Video Customization

Retrieval is Accurate Generation

Measuring and Reducing Model Update Regression in Structured Prediction for NLP

Semi-supervised 3D Object Detection with PatchTeacher and PillarMix

Pseudo Label Refinery for Unsupervised Domain Adaptation on Cross-dataset 3D Object Detection

Towards Fine-Grained HBOE with Rendered Orientation Set and Laplace Smoothing

STraj: Self-training for Bridging the Cross-Geography Gap in Trajectory Prediction

Empowering Self-Learning of LLMs: Inner Knowledge Explicitation as a Catalyst

papers (27)

CLRNet: Cross Layer Refinement Network for Lane Detection

Sparse Fuse Dense: Towards High Quality 3D Detection With Depth Completion

Do Wider Neural Networks Really Help Adversarial Robustness?

Learning to Break the Loop: Analyzing and Mitigating Repetitions for Neural Text Generation

Graph R-CNN: Towards Accurate 3D Object Detection with Semantic-Decorated Local Graph

DID-M3D: Decoupling Instance Depth for Monocular 3D Object Detection

Learning To Affiliate: Mutual Centralized Learning for Few-Shot Classification

Lidar Point Cloud Guided Monocular 3D Object Detection

MonoNeRD: NeRF-like Representations for Monocular 3D Object Detection

One-shot Implicit Animatable Avatars with Model-based Priors

Towards Efficient Adversarial Training on Vision Transformers

Frame-Wise Action Representations for Long Videos via Sequence Contrastive Learning

Learning Occupancy for Monocular 3D Object Detection

Repetition In Repetition Out: Towards Understanding Neural Text Degeneration from the Data Perspective

TagCLIP: A Local-to-Global Framework to Enhance Open-Vocabulary Multi-Label Classification of CLIP without Training

Salient Object Ranking With Position-Preserved Attention

PersonalVideo: High ID-Fidelity Video Customization without Dynamic and Semantic Degradation

3426 Regulating Intermediate 3D Features for Vision-Centric Autonomous Driving

TASeg: Temporal Aggregation Network for LiDAR Semantic Segmentation

MagicID: Hybrid Preference Optimization for ID-Consistent and Dynamic-Preserved Video Customization

Retrieval is Accurate Generation

Measuring and Reducing Model Update Regression in Structured Prediction for NLP

Semi-supervised 3D Object Detection with PatchTeacher and PillarMix

Pseudo Label Refinery for Unsupervised Domain Adaptation on Cross-dataset 3D Object Detection

Towards Fine-Grained HBOE with Rendered Orientation Set and Laplace Smoothing

STraj: Self-training for Bridging the Cross-Geography Gap in Trajectory Prediction

Empowering Self-Learning of LLMs: Inner Knowledge Explicitation as a Catalyst