Video Chaptering on VidChapters-7M Medium (test)
46.7F1 ScoreChapter-Llama
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Chapter-LlamaBackbone=Llama 3.1-8B, Frame selection=Proposed, Finetuned=true2025.03 | 46.7 | 72.3 | 18.8 | 98.7 | |
| Gemini-2.0-FlashBackbone=Gemini-2.0-Flash, Frame selection=Speech-based, Finetuned=false2025.03 | 43.8 | 71.4 | 11.2 | 70.3 | |
| Gemini-1.5-ProBackbone=Gemini-1.5-Pro, Frame selection=Speech-based, Finetuned=false2025.03 | 43.8 | 71.8 | 11.2 | 61.4 | |
| GPT-4oBackbone=GPT-4o, Frame selection=Speech-based, Finetuned=false2025.03 | 38.1 | 68.8 | 8.1 | 51.4 | |
| Llama 3.1-8BBackbone=Llama 3.1-8B, Frame selection=Speech-based, Finetuned=false2025.03 | 30.6 | 62.7 | 5.4 | 28.1 | |
| GPT-4o-miniBackbone=GPT-4o-mini, Frame selection=Speech-based, Finetuned=false2025.03 | 30.5 | 62.3 | 6.1 | 30.6 | |
| Vid2SeqBackbone=Vid2Seq, Frame selection=Equidistant, Finetuned=true2025.03 | 19 | 53.3 | 7.5 | 31.9 | |
| Vid2SeqBackbone=Vid2Seq, Frame selection=Equidistant, Finetuned=false2025.03 | 3.2 | 29.7 | 0.3 | 0.4 |