StreamQA-120K: 1.28 Million Videos for Streaming Video Question Answering

StreamingQA-120K is a large-scale dataset containing 1.28 million videos sourced from WebVid, InternVid, and Panda-70M. It was created by the author 'interlive' and last updated in December 2025. Videos are grouped by similarity for evaluating models on streaming video understanding tasks.

Use Cases

• Train models for video question answering based on grouped video captions.
• Benchmark streaming video understanding performance based on similarity-grouped videos.
• Develop retrieval-augmented generation systems for video content based on video IDs and file paths.
• Fine-tune multimodal models on video-text pairs for caption-based reasoning.

Strengths

• Contains 1.28 million videos, providing a large-scale resource.
• Videos are sourced from three established collections: WebVid, InternVid, and Panda-70M.
• Videos are grouped by similarity, which may facilitate structured evaluation.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and file formats are unknown, which may limit suitability assessment.
• Data may reflect source bias inherent to the three video collections used.

Provenance

Source

WebVid, InternVid, Panda-70M

Collection Method

Videos sourced and grouped by similarity.

Time Range

null

Freshness

Last updated 2025-12-07 06:59:13; freshness should be verified.

Geography

null