Yin Cui

Affiliations

Google

papers

4,139

total citations

papers (16)

Simple Copy-Paste Is a Strong Data Augmentation Method for Instance Segmentation

CVPR 2021arXiv

1,178

citations

Rethinking Pre-training and Self-training

NEURIPS 2020arXiv

712

citations

VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text

NEURIPS 2021arXiv

689

citations

Alternating Gradient Descent and Mixture-of-Experts for Integrated Multimodal Perception

NEURIPS 2023arXiv

citations

Contextualized Spatio-Temporal Contrastive Learning With Self-Supervision

CVPR 2022arXiv

citations

Module-wise Adaptive Distillation for Multimodality Foundation Models

NEURIPS 2023arXiv

citations

Unified Visual Relationship Detection with Vision and Language Models

ICCV 2023arXiv

citations

ArtiScene: Language-Driven Artistic 3D Scene Generation Through Image Intermediary

CVPR 2025arXiv

citations

Train-Once-for-All Personalization

CVPR 2023

citations

Yin Cui

Affiliations

papers (16)

Simple Copy-Paste Is a Strong Data Augmentation Method for Instance Segmentation

Rethinking Pre-training and Self-training

VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text

Spatiotemporal Contrastive Video Representation Learning

Scaling Open-Vocabulary Image Segmentation with Image-Level Labels

SpineNet: Learning Scale-Permuted Backbone for Recognition and Localization

Fashionpedia: Ontology, Segmentation, and an Attribute Localization Dataset

Describe Anything: Detailed Localized Image and Video Captioning

DaTaSeg: Taming a Universal Multi-Dataset Multi-Task Segmentation Model

Visual Fact Checker: Enabling High-Fidelity Detailed Caption Generation

Alternating Gradient Descent and Mixture-of-Experts for Integrated Multimodal Perception

Contextualized Spatio-Temporal Contrastive Learning With Self-Supervision

Module-wise Adaptive Distillation for Multimodality Foundation Models

Unified Visual Relationship Detection with Vision and Language Models

ArtiScene: Language-Driven Artistic 3D Scene Generation Through Image Intermediary

Train-Once-for-All Personalization

papers (16)

Simple Copy-Paste Is a Strong Data Augmentation Method for Instance Segmentation

Rethinking Pre-training and Self-training

VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text

Spatiotemporal Contrastive Video Representation Learning

Scaling Open-Vocabulary Image Segmentation with Image-Level Labels

SpineNet: Learning Scale-Permuted Backbone for Recognition and Localization

Fashionpedia: Ontology, Segmentation, and an Attribute Localization Dataset

Describe Anything: Detailed Localized Image and Video Captioning

DaTaSeg: Taming a Universal Multi-Dataset Multi-Task Segmentation Model

Visual Fact Checker: Enabling High-Fidelity Detailed Caption Generation

Alternating Gradient Descent and Mixture-of-Experts for Integrated Multimodal Perception

Contextualized Spatio-Temporal Contrastive Learning With Self-Supervision

Module-wise Adaptive Distillation for Multimodality Foundation Models

Unified Visual Relationship Detection with Vision and Language Models

ArtiScene: Language-Driven Artistic 3D Scene Generation Through Image Intermediary

Train-Once-for-All Personalization