EgoSchema

EgoSchema is a benchmark for evaluating egocentric video question answering, containing datasets that assess a model's ability to reason and comprehend visual information from first-person video streams.

Papers using EgoSchema (8)

Bridging VideoQA and Video-Guided Agentic Tasks via Generalized Keyframe Extraction2026Building a Mind Palace: Structuring Environment-Grounded Semantic Graphs for Effective Long Video Analysis with LLMs2025 · 1 citesProgressive Video Condensation with MLLM Agent for Long-form Video Understanding2026ReDiPrune: Relevance-Diversity Pre-Projection Token Pruning for Efficient Multimodal LLMs2026CASHEW: Stabilizing Multimodal Reasoning via Iterative Trajectory Aggregation2026Egovlm: Policy Optimization For Egocentric Video Understanding2025VideoAgent: Long-form Video Understanding with Large Language Model as Agent2024 · 42 citesLanguage Repository for Long Video Understanding2024