Object Detection on PDM benchmark leaderboard · SOTA2 Research

Object Detection on PDM

62.1mIoUNo-Time-To-Train

Evaluation Results

Method	Links
No-Time-To-Train2026.07		62.1	65.5
No-Time-To-Train2026.07		55.6	63.6
Qwen3-VL-235B + IPLoc-ID2026.07		53.6	97.6
Qwen3-VL-32B + IPLoc-ID2026.07		50.7	99.5
Qwen3-VL-8B + IPLoc2026.07		48.7	66.7
Qwen3-VL-8B + IPLoc-ID2026.07		48.7	98.7
Qwen3-VL-235B + IPLoc2026.07		48.1	66.6
Qwen3-VL-32B + IPLoc2026.07		47.3	66.7
Qwen3-VL-32B + IPLoc-ID2026.07		45.5	98.6
Qwen3-VL-235B + IPLoc2026.07		44.5	66.8
Qwen3-VL-235B + IPLoc-ID2026.07		44.5	97
Qwen3-VL-32B2026.07		44	65.4
Qwen3-VL-8B + IPLoc-ID2026.07		43.9	94.1
Qwen3-VL-32B + IPLoc2026.07		43.7	66.7
Qwen3-VL-8B + IPLoc2026.07		43.6	66.7
Qwen3-VL-8B2026.07		39.8	91.4
Qwen3-VL-32B2026.07		39.1	69.9
Qwen3-VL-235B2026.07		38.7	63.3
Qwen3-VL-235B2026.07		36.8	77.8
Qwen2-VL-7B + IPLoc2026.07		36.7	66.7
Qwen3-VL-8B2026.07		36.3	86.1
Qwen2-VL-7B + IPLoc-ID2026.07		35.1	98.8
IPLoc 7B (official)2026.07		34.3	71.3
Qwen2-VL-7B + IPLoc2026.07		31.8	66.6
Qwen2-VL-7B + IPLoc-ID2026.07		31.6	97.6
IPLoc 7B (official)2026.07		31.5	72.2
VFM2026.07		28.7	68.1
VFM2026.07		26.4	65.8
Qwen2-VL-7B2026.07		22.9	45.5
Grounding-DINO2026.07		21.3	49.8
Qwen2-VL-7B2026.07		20.4	38
Gemma3-12B + IPLoc-ID2026.07		20.1	95.9
Gemma3-12B + IPLoc2026.07		17.7	66.6
Florence-22026.07		17.4	46.7
Gemma3-12B (+ prompt)2026.07		14.2	66.3
Gemma3-12B + IPLoc2026.07		13.4	66.7
Gemma3-12B + IPLoc-ID2026.07		13.4	93.2
LLaVA1.5-7B + prompt2026.07		12.2	64.4
Gemma3-12B (+ prompt)2026.07		11.7	76
LLaVA1.5-7B + prompt2026.07		0.9	10