Research map / Large language models / Language, safety and interpretability
How2Sign: A Large-Scale Multimodal Dataset for Continuous American Sign Language
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Generate signer videos from keypoints with EDN instead of displaying skeleton visualizations | How2Sign keypoint-based signer video evaluation · ASL signer video understanding | accuracy: 83.3 % -> 91.6 % | Helped |
| Generate signer videos from keypoints with EDN instead of displaying skeleton visualizations | How2Sign keypoint-based signer video evaluation · ASL signer video understanding | BLEU-4: 1.25 -> 1.89 | Helped |
| Use high-resolution (1280 x 720) rather than low-resolution (210 x 260) video for OpenPose extraction | How2Sign 2D pose extraction · How2Sign | average keypoint confidence: 42.4\% confidence for low resolution versus 53.4\% confidence for high resolution | Helped |
| Use high-resolution rather than low-resolution video for OpenPose extraction | How2Sign 2D pose extraction · How2Sign hands | keypoint confidence: 0.24 -> 0.42; 0.30 -> 0.47 | Helped |
| Use generated videos rather than skeleton visualizations | How2Sign keypoint-based signer video evaluation · ASL signer video understanding | BLEU-1: 10.90 -> 12.38 | Helped |
Closest papers by meaning
- Lost in Translation, Found in Embeddings: Sign Language Translation and Alignment (NeurIPS 2026 · similarity 0.91)
- YouTube-SL-25: A Large-Scale, Open-Domain Multilingual Sign Language Parallel Corpus (ICLR 2025 · similarity 0.90)
- Lost in Translation, Found in Context: Sign Language Translation with Contextual Cues (CVPR 2025 · similarity 0.90)
- MM-WLAuslan: Multi-View Multi-Modal Word-Level Australian Sign Language Recognition Dataset (NeurIPS 2024 · similarity 0.89)
- Sign2GPT: Leveraging Large Language Models for Gloss-Free Sign Language Translation (ICLR 2024 · similarity 0.89)
- Scaling Sign Language Translation (NeurIPS 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.