Visual Question Answering on VQA v1
83.2AccuracyGoM (Object Num IDs + Relation labels)
Evaluation Results
| Method | Links | |
|---|---|---|
| GoM (Object Num IDs + Relation labels)MLM=LlamaV-o1-11B2026.03 | 83.2 | |
| GoM (Object Text IDs + Relation labels)MLM=Qwen-2.5-VL-7B2026.03 | 82.5 | |
| GoM (Object Num IDs + Relation labels)MLM=Qwen-2.5-VL-7B2026.03 | 81.1 | |
| GoM (Object Text IDs + Relation labels)MLM=LlamaV-o1-11B2026.03 | 79.8 | |
| GoM (Object Num IDs)MLM=Qwen-2.5-VL-7B2026.03 | 78.1 | |
| GoM (Object Text IDs)MLM=Qwen-2.5-VL-7B2026.03 | 77.9 | |
| GoM (Object Num IDs)MLM=LlamaV-o1-11B2026.03 | 77.9 | |
| Raw imageMLM=Qwen-2.5-VL-7B2026.03 | 77.7 | |
| GoM (Object Text IDs)MLM=LlamaV-o1-11B2026.03 | 76 | |
| Raw imageMLM=LlamaV-o1-11B2026.03 | 75.3 | |
| GoM (Object Text IDs + Relation labels)MLM=Gemma-3-4B2026.03 | 74.2 | |
| SoMMLM=LlamaV-o1-11B2026.03 | 72.8 | |
| Segmented objectsMLM=LlamaV-o1-11B2026.03 | 72.6 | |
| GoM (Object Num IDs)MLM=Gemma-3-4B2026.03 | 71.5 | |
| GoM (Object Num IDs + Relation labels)MLM=Gemma-3-4B2026.03 | 71.2 | |
| SoMMLM=Qwen-2.5-VL-7B2026.03 | 65.4 | |
| GoM (Object Text IDs)MLM=Gemma-3-4B2026.03 | 65.2 | |
| Segmented objectsMLM=Qwen-2.5-VL-7B2026.03 | 65.2 | |
| Raw imageMLM=Gemma-3-4B2026.03 | 64.3 | |
| SoMMLM=Gemma-3-4B2026.03 | 63.8 | |
| Segmented objectsMLM=Gemma-3-4B2026.03 | 63.7 | |
| TVLTInput Modality=Visual, Audio, Pretrain Datasets=YTT-S2022.09 | 58.9 | |
| TVLTInput Modality=Visual, Audio, Pretrain Datasets=HT100M2022.09 | 58.6 | |
| TextModInput Modality=Visual, Text2022.09 | 56.7 | |
| SpeechModInput Modality=Visual, Audio2022.09 | 47 |