Camouflaged Object Detection

We present a comprehensive study on a new task named camouflaged object detection (COD), which aims to identify objects that are "seamlessly" embedded in their surroundings. The high intrinsic similarities between the target object and the background make COD far more challenging than the traditional object detection task... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK USES EXTRA
TRAINING DATA
BENCHMARK
Camouflaged Object Segmentation CAMO SINet* MAE 0.100 # 3
Weighted F-Measure 60.6 # 4
S-Measure 75.1 # 3
E-Measure 77.1 # 4
Camouflaged Object Segmentation CAMO SINet MAE 0.131 # 6
Weighted F-Measure 47.6 # 7
S-Measure 70.8 # 6
E-Measure 70.6 # 6
Camouflaged Object Segmentation COD SINet MAE 0.092 # 4
Weighted F-Measure 35.2 # 5
S-Measure 68.5 # 4
E-Measure 71.8 # 4
Camouflaged Object Segmentation COD SINet* MAE 0.051 # 1
Weighted F-Measure 55.1 # 1
S-Measure 77.1 # 1
E-Measure 80.6 # 1

Methods used in the Paper