CornerNet-Lite: Efficient Keypoint Based Object Detection

18 Apr 2019Hei LawYun TengOlga RussakovskyJia Deng

Keypoint-based methods are a relatively new paradigm in object detection, eliminating the need for anchor boxes and offering a simplified detection framework. Keypoint-based CornerNet achieves state of the art accuracy among single-stage detectors... (read more)

PDF Abstract

Evaluation Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK COMPARE
Real-Time Object Detection COCO CornerNet-Squeeze MAP 34.4 # 5
Real-Time Object Detection COCO CornerNet-Squeeze FPS 33 # 9
Object Detection COCO minival CornerNet-Saccade (Hourglass-104) box AP 41.4 # 29
Object Detection COCO minival CornerNet-Saccade (Hourglass-104) APS 23.8 # 19
Object Detection COCO minival CornerNet-Saccade (Hourglass-104) APM 43.5 # 24
Object Detection COCO minival CornerNet-Saccade (Hourglass-104) APL 57.1 # 12
Object Detection COCO minival CornerNet-Saccade (Hourglass-54) box AP 42.6 # 21
Object Detection COCO minival CornerNet-Saccade (Hourglass-54) APS 25.5 # 15
Object Detection COCO minival CornerNet-Saccade (Hourglass-54) APM 44.3 # 19
Object Detection COCO minival CornerNet-Saccade (Hourglass-54) APL 58.4 # 8
Object Detection COCO test-dev CornerNet-Saccade (Hourglass-104, multi-scale) box AP 43.2 # 28
Object Detection COCO test-dev CornerNet-Saccade (Hourglass-104, multi-scale) APS 24.4 # 31
Object Detection COCO test-dev CornerNet-Saccade (Hourglass-104, multi-scale) APM 44.6 # 31
Object Detection COCO test-dev CornerNet-Saccade (Hourglass-104, multi-scale) APL 57.3 # 13
Object Detection COCO test-dev CornerNet-Squeeze box AP 34.4 # 69