paper-with-me

Zero-Shot Video Question Answer 벤치마크

Zero-Shot Video Question Answer on MSRVTT-QA

60개 결과 · ⬇ CSV · JSON

Accuracy

29.6 40.3 51 61.7 72.4 2023-04 2026-11 LLaMA Adapter-7B — 43.8 (2023-04-28) LLaMA Adapter-7B — 43.8 (2023-04-28) Video Chat-7B — 45.0 (2023-05-10) Video Chat-7B — 45.0 (2023-05-10) Video LLaMA-7B — 29.6 (2023-06-05) Video LLaMA-7B — 29.6 (2023-06-05) Video-ChatGPT-7B — 49.3 (2023-06-08) Video-ChatGPT-7B — 49.3 (2023-06-08) MovieChat — 52.7 (2023-07-31) MovieChat — 52.7 (2023-07-31) Omni-VideoAssistant — 55.3 (2023-08-08) Omni-VideoAssistant — 55.3 (2023-08-08) BT-Adapter (zero-shot) — 51.2 (2023-09-27) BT-Adapter (zero-shot) — 51.2 (2023-09-27) BT-Adapter (zero-shot) — 51.2 (2023-09-27) BT-Adapter (zero-shot) — 51.2 (2023-09-27) Chat-UniVi-7B — 55.0 (2023-11-14) Chat-UniVi-7B — 55.0 (2023-11-14) Video-LLaVA-7B — 59.2 (2023-11-16) Video-LLaVA-7B — 59.2 (2023-11-16) LLaMA-VID-13B (2 Token) — 58.9 (2023-11-28) LLaMA-VID-7B (2 Token) — 57.7 (2023-11-28) VideoChat2 — 54.1 (2023-11-28) LLaMA-VID-13B (2 Token) — 58.9 (2023-11-28) LLaMA-VID-7B (2 Token) — 57.7 (2023-11-28) VideoChat2 — 54.1 (2023-11-28) Vista-LLaMA-7B — 60.5 (2023-12-12) Vista-LLaMA-7B — 60.5 (2023-12-12) SUM-shot+Vicuna — 56.8 (2023-12-16) SUM-shot+Vicuna — 56.8 (2023-12-16) Video-LaVIT — 59.3 (2024-02-05) Video-LaVIT — 59.3 (2024-02-05) CAT-7B — 62.1 (2024-03-07) CAT-7B — 62.1 (2024-03-07) Elysium — 67.5 (2024-03-25) Elysium — 67.5 (2024-03-25) IG-VLM — 63.8 (2024-03-27) IG-VLM — 63.8 (2024-03-27) ST-LLM — 63.2 (2024-03-30) ST-LLM — 63.2 (2024-03-30) MiniGPT4-video-7B — 59.73 (2024-04-04) MiniGPT4-video-7B — 59.73 (2024-04-04) PLLaVA (34B) — 68.7 (2024-04-25) PLLaVA (34B) — 68.7 (2024-04-25) Flash-VStream — 72.4 (2024-06-12) Flash-VStream — 72.4 (2024-06-12) VideoGPT+ — 60.6 (2024-06-13) VideoGPT+ — 60.6 (2024-06-13) Tarsier (34B) — 66.4 (2024-06-30) Tarsier (34B) — 66.4 (2024-06-30) SlowFast-LLaVA-34B — 67.4 (2024-07-22) SlowFast-LLaVA-34B — 67.4 (2024-07-22) PPLLaVA-7B — 64.3 (2024-11-04) PPLLaVA-7B — 64.3 (2024-11-04) TS-LLaVA-34B — 66.2 (2024-11-17) TS-LLaVA-34B — 66.2 (2024-11-17) LinVT-Qwen2-VL (7B) — 66.2 (2024-12-06) LinVT-Qwen2-VL (7B) — 66.2 (2024-12-06) LLaVA-Mini — 59.5 (2025-01-07) LLaVA-Mini — 59.5 (2025-01-07) LLaMA Adapter-7B — 43.8 (2023-04-28) Video Chat-7B — 45.0 (2023-05-10) Video-ChatGPT-7B — 49.3 (2023-06-08) MovieChat — 52.7 (2023-07-31) Omni-VideoAssistant — 55.3 (2023-08-08) Video-LLaVA-7B — 59.2 (2023-11-16) Vista-LLaMA-7B — 60.5 (2023-12-12) CAT-7B — 62.1 (2024-03-07) Elysium — 67.5 (2024-03-25) PLLaVA (34B) — 68.7 (2024-04-25) Flash-VStream — 72.4 (2024-06-12)
RankModel AccuracyConfidence Score Extra Training Data PaperCodeYear
1 Flash-VStream 72.43.4 Flash-VStream: Memory-Based Real-Time Understanding for Long Video Streams IVGSZ/Flash-VStream 2024
2 PLLaVA (34B) 68.73.6 PLLaVA : Parameter-free LLaVA Extension from Images to Videos for Video Dense Captioning magic-research/PLLaVA 2024
3 Elysium 67.53.2 Elysium: Exploring Object-level Perception in Videos via MLLM hon-wong/elysium 2024
4 SlowFast-LLaVA-34B 67.43.7 SlowFast-LLaVA: A Strong Training-Free Baseline for Video Large Language Models apple/ml-slowfast-llava 2024
5 Tarsier (34B) 66.43.7 Tarsier: Recipes for Training and Evaluating Large Video Description Models bytedance/tarsier 2024
6 LinVT-Qwen2-VL (7B) 66.24.0 LinVT: Empower Your Image-level Large Language Model to Understand Videos gls0425/linvt 2024
6 TS-LLaVA-34B 66.23.6 TS-LLaVA: Constructing Visual Tokens through Thumbnail-and-Sampling for Training-Free Video Large Language Models tingyu215/ts-llava 2024
8 PPLLaVA-7B 64.33.5 PPLLaVA: Varied Video Sequence Understanding With Prompt Guidance farewellthree/ppllava 2024
9 IG-VLM 63.83.5 An Image Grid Can Be Worth a Video: Zero-shot Video Question Answering Using a VLM imagegridworth/IG-VLM 2024
10 ST-LLM 63.23.4 ST-LLM: Large Language Models Are Effective Temporal Learners TencentARC/ST-LLM 2024
11 CAT-7B 62.13.5 CAT: Enhancing Multimodal Large Language Model to Answer Questions in Dynamic Audio-Visual Scenarios rikeilong/bay-cat 2024
12 VideoGPT+ 60.63.6 VideoGPT+: Integrating Image and Video Encoders for Enhanced Video Understanding mbzuai-oryx/videogpt-plus 2024
13 Vista-LLaMA-7B 60.53.3 Vista-LLaMA: Reliable Video Narrator via Equal Distance to Visual Tokens 2023
14 MiniGPT4-video-7B 59.73– MiniGPT4-Video: Advancing Multimodal LLMs for Video Understanding with Interleaved Visual-Textual Tokens Vision-CAIR/MiniGPT4-video · pwc-1/Paper-9 2024
15 LLaVA-Mini 59.53.6 LLaVA-Mini: Efficient Image and Video Large Multimodal Models with One Vision Token ictnlp/llava-mini 2025
16 Video-LaVIT 59.33.3 Video-LaVIT: Unified Video-Language Pre-training with Decoupled Visual-Motional Tokenization jy0205/lavit 2024
17 Video-LLaVA-7B 59.23.5 ✓ Video-LLaVA: Learning United Visual Representation by Alignment Before Projection PKU-YuanGroup/Video-LLaVA · PKU-YuanGroup/MoE-LLaVA · pku-yuangroup/languagebind · +3 2023
18 LLaMA-VID-13B (2 Token) 58.93.3 LLaMA-VID: An Image is Worth 2 Tokens in Large Language Models lastmile-ai/aiconfig · dvlab-research/llama-vid 2023
19 LLaMA-VID-7B (2 Token) 57.73.2 LLaMA-VID: An Image is Worth 2 Tokens in Large Language Models lastmile-ai/aiconfig · dvlab-research/llama-vid 2023
20 SUM-shot+Vicuna 56.8– Shot2Story20K: A New Benchmark for Comprehensive Understanding of Multi-shot Videos bytedance/Shot2Story 2023
1–20 / 60 다음 → 페이지당 10 20 50 100