Research map / Detection and segmentation / Open-set and few-shot detection
Contextual Object Detection with Multimodal Large Language Models
Yuhang Zang, Wei Li, Jun Han, Kaiyang Zhou, Chen Change Loy
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add local visual tokens as LLM prefix (ablation) | ContextDET · CODE val cloze test | Acc@1: 30.9 -> 48.7 | Improved |
| add local visual tokens as LLM prefix (ablation) | ContextDET · CODE val cloze test | AP@1: 4.0 -> 10.4 | Improved |
| use p=9 local tokens instead of 4 (ablation) | ContextDET · CODE val cloze test | AP@1: 10.1 -> 10.4 | Improved |
| use p=9 local tokens instead of 16 (ablation) | ContextDET · CODE val cloze test | AP@1: 9.9 -> 10.4 | Improved |
| replace ResNet50 with Swin-B backbone (ablation) | ContextDET (OPT-2.7B) · CODE val cloze test | AP@1: 10.2 -> 13.1 | Improved |
| replace OPT-2.7B with OPT-6.7B (Swin-B) (ablation) | ContextDET · CODE val cloze test | AP@1: 13.1 -> 13.7 | Improved |
| replace OPT-2.7B with OPT-6.7B (ResNet50) (ablation) | ContextDET · CODE val cloze test | AP@1: 10.2 -> 11.1 | Improved |
| end-to-end ContextDET instead of cascade LLaVA1.5+GLIP | ContextDET · CODE val cloze test | Acc@1: 42.9 -> 48.7 | Improved |
| fine-tune GLIP on CODE in cascade | LLaVA1.5+GLIP · CODE val cloze test | AP@1: 9.2 -> 9.1 | No clear effect |
| use ContextDET instead of BARON on OV-COCO | ContextDET · OV-COCO (novel classes) | AP50 novel: 34.0 -> 36.8 | Improved |
| use ContextDET instead of PolyFormer | ContextDET · RefCOCO val | mIoU: 75.96 -> 76.40 | Improved |
| use ContextDET instead of Deformable DETR | ContextDET · COCO val | AP: 46.2 -> 43.4 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- Deformable DETR — Deformable DETR: Deformable Transformers for End-to-End Object Detection (ICLR 2021) · better on 0 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- ROD-MLLM: Towards More Reliable Object Detection in Multimodal Large Language Models (CVPR 2025 · similarity 0.90)
- On Scaling Up a Multilingual Vision and Language Model (CVPR 2024 · similarity 0.89)
- ConTextual: Evaluating Context-Sensitive Text-Rich Visual Reasoning in Large Multimodal Models (ICML 2024 · similarity 0.88)
- ORIC: Benchmarking Object Recognition under Contextual Incongruity in Large Vision-Language Models (CVPR 2026 · similarity 0.88)
- Improving Context Understanding in Multimodal Large Language Models via Multimodal Composition Learning (ICML 2024 · similarity 0.88)
- LLMDet: Learning Strong Open-Vocabulary Object Detectors under the Supervision of Large Language Models (CVPR 2025 · similarity 0.88)
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.
