YOLO9000: Better, Faster, Stronger

CVPR 2017 Joseph RedmonAli Farhadi

We introduce YOLO9000, a state-of-the-art, real-time object detection system that can detect over 9000 object categories. First we propose various improvements to the YOLO detection method, both novel and drawn from prior work... (read more)

PDF Abstract

Code


Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK USES EXTRA
TRAINING DATA
RESULT LEADERBOARD
Object Detection COCO test-dev YOLO v2 + Darknet-19 box AP 21.6 # 83
Object Detection PASCAL VOC 2007 YOLO v2 MAP 78.6% # 12

Results from Other Papers


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK USES EXTRA
TRAINING DATA
SOURCE PAPER COMPARE
Object Counting CARPK YOLO9000opt (2017) MAE 130.40 # 7
RMSE 172.46 # 7
Dense Object Detection SKU-110K YOLO9000opt AP .094 # 4
AP75 .073 # 3

Methods used in the Paper