Feng Cheng

10

Papers

17

Total Citations

Papers (10)

Synthetic Video Enhances Physical Fidelity in Video Synthesis

TaskGalaxy: Scaling Multi-modal Instruction Fine-tuning with Tens of Thousands Vision Task Types

VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos

VideoAuteur: Towards Long Narrative Video Generation

LoCoNet: Long-Short Context Network for Active Speaker Detection

Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives

Stochastic Backpropagation: A Memory Efficient Strategy for Training Video Models

VindLU: A Recipe for Effective Video-and-Language Pretraining

Unified Coarse-to-Fine Alignment for Video-Text Retrieval

TALLFormer: Temporal Action Localization with a Long-Memory Transformer