List your company

Object Counting on LVIS-Count 100-image zero-shot (test)

23.36MAEVTOS Analyzer ON
22.12430.46738.8147.153Jun 17, 2026

Evaluation results

MethodLinks
VTOS Analyzer ONFamily=VTOS (ours), Analyzer=ON, LLM=Sonnet 4.6, VFM=Grounding DINO + SAM2026.0623.3637.774.0240.48
VTOS Analyzer OFFFamily=VTOS (ours), Analyzer=OFF, LLM=Sonnet 4.6, VFM=Grounding DINO + SAM2026.0632.8347.8616.3936.75
Sonnet 4.6 (Count Only)Family=End-to-End VLM, Mode=Count Only2026.0635.1356.884.35—
Sonnet 4.6 (Direct Bbox)Family=End-to-End VLM, Mode=Direct Bbox2026.0635.1842.3133.647.54
CodeAct (Sonnet 4.6 + tools)Family=Agentic, Components=Sonnet 4.6 + tools2026.0639.3849.6632.429.21
CountGD (vanilla)Family=Specialist, Variant=vanilla2026.0643.01126.8232.0536.13
Reflexion (Sonnet 4.6 + visual critic)Family=Agentic, Components=Sonnet 4.6 + visual critic2026.0644.2856.5425.8824.91
ReAct (Sonnet 4.6 + DINO)Family=Agentic, Components=Sonnet 4.6 + DINO2026.0646.3152.646.0319.66
SAHI (DINO)Family=Heuristics, Base Model=DINO2026.0646.5750.4244.6515.91
Qwen3-VL-32B (Direct)Family=End-to-End VLM, Mode=Direct, Model Size=32B2026.0648.3247.2647.269.98
CountGD (Crop+NMS)Family=Specialist, Variant=Crop+NMS2026.0648.84156.1342.5236.59
Grounding DINO (thr=0.10)Family=Specialist, Threshold=0.102026.0652.7677.6127.4432.19
Qwen3-VL-8B (Direct)Family=End-to-End VLM, Mode=Direct, Model Size=8B2026.0654.2654.2654.263.38