Xiangxiang Chu

Affiliations

Alibaba

papers

2,907

total citations

papers (20)

Twins: Revisiting the Design of Spatial Attention in Vision Transformers

NEURIPS 2021arXiv

1,237

citations

FairNAS: Rethinking Evaluation Fairness of Weight Sharing Neural Architecture Search

ICCV 2021arXiv

357

citations

SCTNet: Single Branch CNN with Transformer Semantic Information for Real-Time Segmentation

AAAI 2024arXiv

130

citations

PeLK: Parameter-efficient Large Kernel ConvNets with Peripheral Convolution

CVPR 2024arXiv

citations

Norm Tweaking: High-Performance Low-Bit Quantization of Large Language Models

AAAI 2024arXiv

citations

LiDAR-PTQ: Post-Training Quantization for Point Cloud 3D Object Detection

ICLR 2024arXiv

citations

Modeling Motion With Multi-Modal Features for Text-Based Video Segmentation

CVPR 2022arXiv

citations

AeDet: Azimuth-Invariant Multi-View 3D Object Detection

CVPR 2023arXiv

citations

MixPath: A Unified Approach for One-shot Neural Architecture Search

ICCV 2023arXiv

citations

VMBench: A Benchmark for Perception-Aligned Video Motion Generation

ICCV 2025arXiv

citations

USP: Unified Self-Supervised Pretraining for Image Generation and Understanding

ICCV 2025arXiv

citations

Revealing the Dark Secrets of Extremely Large Kernel ConvNets on Robustness

ICML 2024arXiv

citations

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

ICCV 2023arXiv

citations

LD-RPS: Zero-Shot Unified Image Restoration via Latent Diffusion Recurrent Posterior Sampling

ICCV 2025arXiv

citations

Towards Efficient Foundation Model for Zero-shot Amodal Segmentation

CVPR 2025

citations

UPRE: Zero-Shot Domain Adaptation for Object Detection via Unified Prompt and Representation Enhancement

ICCV 2025arXiv

citations

Xiangxiang Chu

Affiliations

papers (20)

Twins: Revisiting the Design of Spatial Attention in Vision Transformers

FairNAS: Rethinking Evaluation Fairness of Weight Sharing Neural Architecture Search

Fair DARTS: Eliminating Unfair Advantages in Differentiable Architecture Search

PromptDet: Towards Open-Vocabulary Detection Using Uncurated Images

SegViT: Semantic Segmentation with Plain Vision Transformers

Fully Convolutional One-Stage 3D Object Detection on LiDAR Range Images

SCTNet: Single Branch CNN with Transformer Semantic Information for Real-Time Segmentation

PeLK: Parameter-efficient Large Kernel ConvNets with Peripheral Convolution

Norm Tweaking: High-Performance Low-Bit Quantization of Large Language Models

LiDAR-PTQ: Post-Training Quantization for Point Cloud 3D Object Detection

Modeling Motion With Multi-Modal Features for Text-Based Video Segmentation

AeDet: Azimuth-Invariant Multi-View 3D Object Detection

MixPath: A Unified Approach for One-shot Neural Architecture Search

VMBench: A Benchmark for Perception-Aligned Video Motion Generation

USP: Unified Self-Supervised Pretraining for Image Generation and Understanding

Revealing the Dark Secrets of Extremely Large Kernel ConvNets on Robustness

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

LD-RPS: Zero-Shot Unified Image Restoration via Latent Diffusion Recurrent Posterior Sampling

Towards Efficient Foundation Model for Zero-shot Amodal Segmentation

UPRE: Zero-Shot Domain Adaptation for Object Detection via Unified Prompt and Representation Enhancement

papers (20)

Twins: Revisiting the Design of Spatial Attention in Vision Transformers

FairNAS: Rethinking Evaluation Fairness of Weight Sharing Neural Architecture Search

Fair DARTS: Eliminating Unfair Advantages in Differentiable Architecture Search

PromptDet: Towards Open-Vocabulary Detection Using Uncurated Images

SegViT: Semantic Segmentation with Plain Vision Transformers

Fully Convolutional One-Stage 3D Object Detection on LiDAR Range Images

SCTNet: Single Branch CNN with Transformer Semantic Information for Real-Time Segmentation

PeLK: Parameter-efficient Large Kernel ConvNets with Peripheral Convolution

Norm Tweaking: High-Performance Low-Bit Quantization of Large Language Models

LiDAR-PTQ: Post-Training Quantization for Point Cloud 3D Object Detection

Modeling Motion With Multi-Modal Features for Text-Based Video Segmentation

AeDet: Azimuth-Invariant Multi-View 3D Object Detection

MixPath: A Unified Approach for One-shot Neural Architecture Search

VMBench: A Benchmark for Perception-Aligned Video Motion Generation

USP: Unified Self-Supervised Pretraining for Image Generation and Understanding

Revealing the Dark Secrets of Extremely Large Kernel ConvNets on Robustness

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

LD-RPS: Zero-Shot Unified Image Restoration via Latent Diffusion Recurrent Posterior Sampling

Towards Efficient Foundation Model for Zero-shot Amodal Segmentation

UPRE: Zero-Shot Domain Adaptation for Object Detection via Unified Prompt and Representation Enhancement