Test photo shepherd-sheep

shepherd-sheep

License: CC0-1.0
Author: Mirzoulugʻbek
Source: Wikimedia Commons, https://commons.wikimedia.org/wiki/File:Shepherd_with_a_flock_of_sheep_in_Kyrgyzstan_01.jpg (CC0 1.0 by the author, checked 2026-10-27); downscaled to 1600x1205, metadata removed
Hard cases: crowd, small-objects

Hover any output to magnify the same spot in every output of that task.

Object detection

D-FINE l — D-FINE lFails: labels the black lamb as a dog

D-FINE m

D-FINE n — D-FINE nFails: labels the sheep as horses and a dog

D-FINE s — D-FINE sFails: labels the black lamb as a dog

D-FINE x

DEIMv2 l

DEIMv2 m

DEIMv2 s — DEIMv2 sFails: labels the black lamb as a dog

DEIMv2 x

Grounding DINO base

Grounding DINO tiny — Grounding DINO tinyFails: misses the whole flock of sheep

GTR l

GTR m

GTR s

GTR x

LLMDet base — LLMDet baseFails: misses the whole flock of sheep

LLMDet large — LLMDet largeFails: misses the whole flock of sheep

LLMDet tiny — LLMDet tinyFails: misses the whole flock of sheep

MM-Grounding-DINO base — MM-Grounding-DINO baseFails: misses rider and flock, calls lamb a dog

MM-Grounding-DINO large — MM-Grounding-DINO largeFails: misses the rider and the whole flock

MM-Grounding-DINO tiny — MM-Grounding-DINO tinyFails: misses the rider and the whole flock

RF-DETR large — RF-DETR largeFails: misses the whole flock of sheep

RF-DETR medium — RF-DETR mediumFails: misses the whole flock of sheep

RF-DETR nano

RF-DETR small

RT-DETRv4 l

RT-DETRv4 m

RT-DETRv4 s

RT-DETRv4 x

YOLO11 l

YOLO11 m

YOLO11 n — YOLO11 nFails: misses the whole flock of sheep

YOLO11 s — YOLO11 sFails: misses the man riding the horse

YOLO11 x

YOLO26 l

YOLO26 m

YOLO26 n

YOLO26 s

YOLO26 x

YOLOv8 l

YOLOv8 m

YOLOv8 n

YOLOv8 s

YOLOv8 x

Instance segmentation

EoMT large-1280 — EoMT large-1280Fails: one mask covers the whole back flock

EoMT large-640

GTR-Seg l

GTR-Seg m

GTR-Seg s

GTR-Seg x

RF-DETR-Seg 2xlarge — RF-DETR-Seg 2xlargeFails: misses the whole flock of sheep

RF-DETR-Seg large — RF-DETR-Seg largeFails: misses nearly all of the flock

RF-DETR-Seg medium — RF-DETR-Seg mediumFails: misses the flock, two sheep called dogs

RF-DETR-Seg nano — RF-DETR-Seg nanoFails: misses the whole flock of sheep

RF-DETR-Seg small — RF-DETR-Seg smallFails: misses the whole flock of sheep

RF-DETR-Seg xlarge — RF-DETR-Seg xlargeFails: misses the whole flock of sheep

YOLO11-seg l — YOLO11-seg lFails: black lamb masked as a dog

YOLO11-seg m

YOLO11-seg n — YOLO11-seg nFails: misses nearly the whole flock of sheep

YOLO11-seg s

YOLO11-seg x

YOLO26-seg l

YOLO26-seg m

YOLO26-seg n — YOLO26-seg nFails: misses the whole flock of sheep

YOLO26-seg s

YOLO26-seg x

YOLOv8-seg l — YOLOv8-seg lFails: black lamb masked as a dog

YOLOv8-seg m — YOLOv8-seg mFails: misses nearly the whole flock, black lamb masked as a dog

YOLOv8-seg n — YOLOv8-seg nFails: misses nearly the whole flock of sheep

YOLOv8-seg s — YOLOv8-seg sFails: misses nearly the whole flock, black lamb masked as a dog

YOLOv8-seg x

Pose estimation