Back to datasets
Dataset assetOpen Source CommunityLong Video AnalysisMultimodal Perception

LongVALE

LongVALE: Time-Aware Omni-Modal Perception Benchmark for Vision-Audio-Language-Event in Long-duration Videos

Source
github
Created
Dec 6, 2024
Updated
Dec 6, 2024
Signals
177 views
Availability
Linked source ready
Overview

Dataset description and usage context

LongVALE

Dataset Overview

  • Name: LongVALE
  • Full Name: Vision-Audio-Language-Event Benchmark Towards Time-Aware Omni-Modal Perception of Long Videos
  • Description: This dataset aims to provide a time-aware omni-modal perception benchmark for long-duration videos, covering visual, audio, language, and event modalities.
Need downstream help?

Pair the dataset with AI analysis and content workflows.

Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.

Explore AI studio