Crop Recognition on CDDMBench
97.4AccuracyQwen-VL-Chat-AG (7B)
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen-VL-Chat-AG (7B)Method=SFT (Unfrozen encoder)2026.01 | 97.4 | |
| Qwen2.5-VL-3B-InstructMethod=Reasoning-Enhanced GRPO2026.01 | 92.58 | |
| Qwen2.5-VL-3B-InstructMethod=GRPO2026.01 | 92.33 | |
| Qwen2.5-VL-3B-InstructMethod=SFT2026.01 | 90.97 | |
| Qwen-VL-Chat-AG* (7B)Method=SFT (Frozen encoder)2026.01 | 84.4 | |
| Gpt-5-NanoMethod=+Judge2026.01 | 63.38 | |
| Agri-CPJ (+ LLM-as-a-Judge)Model=GPT-5-Nano, Caption Generator=GPT-5-mini2026.04 | 63.38 | |
| Gpt-5-NanoMethod=Expl. Caption2026.01 | 60.3 | |
| Agri-CPJ (+ Caption (Optimized))Model=GPT-5-Nano, Caption Generator=GPT-5-mini2026.04 | 60.3 | |
| Gpt-5-NanoMethod=+Few-shot2026.01 | 58.9 | |
| Agri-CPJ (+ Few-shot)Model=GPT-5-Nano, Caption Generator=GPT-5-mini2026.04 | 58.9 | |
| Qwen-VL-Chat (7B)Method=+Judge2026.01 | 54.9 | |
| Agri-CPJ (+ LLM-as-a-Judge)Model=Qwen-VL-Chat, Caption Generator=Qwen2.5-VL-72B-Instruct2026.04 | 54.9 | |
| Qwen-VL-Chat (7B)Method=+Few-shot2026.01 | 53.39 | |
| Agri-CPJ (+ Few-shot)Model=Qwen-VL-Chat, Caption Generator=Qwen2.5-VL-72B-Instruct2026.04 | 53.39 | |
| Agri-CPJ (+ LLM-as-a-Judge)Model=GPT-5-Nano, Caption Generator=Qwen2.5-VL-72B-Instruct2026.04 | 53.2 | |
| Agri-CPJ (+ LLM-as-a-Judge)Model=Qwen-VL-Chat, Caption Generator=GPT-5-mini2026.04 | 51.6 | |
| Agri-CPJ (+ Caption (Optimized))Model=GPT-5-Nano, Caption Generator=Qwen2.5-VL-72B-Instruct2026.04 | 51.2 | |
| Agri-CPJ (+ Few-shot)Model=Qwen-VL-Chat, Caption Generator=GPT-5-mini2026.04 | 51 | |
| Agri-CPJ (+ Few-shot)Model=GPT-5-Nano, Caption Generator=Qwen2.5-VL-72B-Instruct2026.04 | 50.1 | |
| Gpt-5-NanoMethod=Zero-shot2026.01 | 47 | |
| Zero-shot (Our Baseline)Model=GPT-5-Nano, Caption Generator=Qwen2.5-VL-72B-Instruct2026.04 | 47 | |
| Zero-shot (Our Baseline)Model=GPT-5-Nano, Caption Generator=GPT-5-mini2026.04 | 47 | |
| Qwen2.5-VL-3B-InstructMethod=Few-shot2026.01 | 36.56 | |
| Qwen-VL-Chat (7B)Method=Expl. Caption2026.01 | 29.3 | |
| Agri-CPJ (+ Caption (Optimized))Model=Qwen-VL-Chat, Caption Generator=Qwen2.5-VL-72B-Instruct2026.04 | 29.3 | |
| Agri-CPJ (+ Caption (Optimized))Model=Qwen-VL-Chat, Caption Generator=GPT-5-mini2026.04 | 28.99 | |
| Zero-shot (Our Baseline)Model=Qwen-VL-Chat, Caption Generator=Qwen2.5-VL-72B-Instruct2026.04 | 28.55 | |
| Qwen2.5-VL-3B-InstructMethod=Zero-shot2026.01 | 28.41 | |
| Qwen-VL-Chat (7B)Method=Zero-shot2026.01 | 28.4 | |
| Zero-shot (Liu et al., 2024) BaselineModel=Qwen-VL-Chat, Caption Generator=Qwen2.5-VL-72B-Instruct2026.04 | 28.4 |