Research map / Vision-language and multimodal / Multimodal LLMs
Multi-label Instance-level Generalised Visual Grounding in Agriculture
Mohammadreza Haghighat, Alzayat Saleh, Mostafa Azghadi
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 word-level cues to sentence-level similarity (ablation) | Weed-VG · gRef-CW Test | Top-1: 49.29 -> 62.42 | Improved |
| add sentence-level similarity to word-level cues (ablation) | Weed-VG · gRef-CW Test | Top-1: 36.19 -> 62.42 | Improved |
| add query projection (ablation) | Weed-VG · gRef-CW Test | Top-1: 33.20 -> 62.42 | Improved |
| add query projection (ablation) | Weed-VG · gRef-CW Test | mIoU: 35.39 -> 57.25 | Improved |
| add hierarchical constraint (ablation) | Weed-VG · gRef-CW Test | Neg-Acc: 41.60 -> 78.35 | Improved |
| add hierarchical constraint (ablation) | Weed-VG · gRef-CW Test | R@0.5: 53.87 -> 55.44 | No clear effect |
| add InterpIoU with distance and size aware matching (ablation) | Weed-VG · gRef-CW Test | mIoU: 47.72 -> 57.25 | Improved |
| add InterpIoU with distance and size aware matching (ablation) | Weed-VG · gRef-CW Test | Top-1: 49.15 -> 62.42 | Improved |
| add InterpIoU with distance and size aware matching (ablation) | Weed-VG · gRef-CW Test | Neg-Acc: 75.88 -> 78.35 | Improved |
| wrap GroundingDINO with HRS (Weed-VG) instead of GDINO-L alone | GroundingDINO · gRef-CW Test | Top-1: 20.38 -> 62.42 | Improved |
| use Weed-VG instead of SAM3 on dense crop scenes (>30 instances) | Weed-VG · gRef-CW Test, crop >30 instances | mIoU: 25.81 -> 47.58 | Improved |
| use Weed-VG with GDINO-L on large crops | Weed-VG · gRef-CW Test, large crops | Top-1: 74.33 -> 71.90 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MDETR — MDETR - Modulated Detection for End-to-End Multi-Modal Understanding (ICCV 2021) · better on 5 of 5 shared results
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
- AgroOmni: A Large-Scale Multi-view Agricultural Dataset for Cross-Scale Multimodal Reasoning (NeurIPS 2026 · similarity 0.89)
- AgMMU: A Comprehensive Agricultural Multimodal Understanding Benchmark (NeurIPS 2025 · similarity 0.88)
- Cropper: Vision-Language Model for Image Cropping through In-Context Learning (CVPR 2025 · similarity 0.87)
- Connecting the Dots: Training-Free Visual Grounding via Agentic Reasoning (AAAI 2026 · similarity 0.86)
- GroundingME: Exposing the Visual Grounding Gap in MLLMs through Multi-Dimensional Evaluation (CVPR 2026 · similarity 0.86)
- GroundVLP: Harnessing Zero-Shot Visual Grounding from Vision-Language Pre-training and Open-Vocabulary Object Detection (AAAI 2024 · similarity 0.86)
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.
