HackerPapers

arXiv:1612.08242cs.CV

YOLO9000: Better, Faster, Stronger

Joseph Redmon, Ali Farhadi

Abstract

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. The improved model, YOLOv2, is state-of-the-art on standard detection tasks like PASCAL VOC and COCO. At 67 FPS, YOLOv2 gets 76.8 mAP on VOC 2007. At 40 FPS, YOLOv2 gets 78.6 mAP, outperforming state-of-the-art methods like Faster RCNN with ResNet and SSD while still running significantly faster. Finally we propose a method to jointly train on object detection and classification. Using this method we train YOLO9000 simultaneously on the COCO detection dataset and the ImageNet classification dataset. Our joint training allows YOLO9000 to predict detections for object classes that don't have labelled detection data. We validate our approach on the ImageNet detection task. YOLO9000 gets 19.7 mAP on the ImageNet detection validation set despite only having detection data for 44 of the 200 classes. On the 156 classes not in COCO, YOLO9000 gets 16.0 mAP. But YOLO can detect more than just 200 classes; it predicts detections for more than 9000 different object categories. And it still runs in real-time.

arXiv comments
Journal reference
DOI
Subjects
cs.CVComputer Vision and Pattern Recognition

On Hacker News

Linked by 5 identifiable commenters in 6 comments across 6 threads, from Jun 2017 to Mar 2020. Found in 2 HN submissions from Dec 2016 to Jul 2020. Open submission discussions.

Comments linking this paper, per year.

Discussed in comments6 threads

  1. The End of Starsky Robotics 1 comment

    • For others like me who were as amused by the name, YOLO9000, it's a real thing, "a state-of-the-art, real-time object detection system that can detect over 9000 object categories". YOLO9000: Better, Faster, Stronger - https://arxiv.org/abs/1612.08242 (2016)

      lioetersView on HNParent context

  2. An intuitive explanation of the SSD MultiBox object detection technique 1 comment

    • The original YOLO is faster (especially the "Fast YOLO" variant), but is a lot less accurate in its detections: Fast YOLO has 52.7 mean Average Precision @155 FPS vs SSD-300 with 74.3 mAP @59 FPS. But the new YOLO9000[1] architecture seems to be even better than SSD! Would like…

      edshiroView on HNParent context

  3. Using CNNs for Handbag Brand and Color Detection 1 comment

    • There are quite a few papers that discuss the application R-CNN (Region-based Convolutional Neural Network) for object detection and classification [1-3]. On websites such as Medium, you can find lots of blog posts on object detection using popular frameworks such as Tensorflow…

      edshiroView on HNParent context

  4. YOLO: Real-Time Object Detection 1 comment

    • This is not "modern" though, the paper is quite old. He has improved it significantly since then: https://arxiv.org/abs/1612.08242. Best paper honorable mention at CVPR2017.

      0xbearView on HNParent context

  5. CVPR 2017 Best Paper Awards 1 comment

    • * Densely Connected Convolutional Networks [0] by Gao Huang, Zhuang Liu, Laurens van der Maaten, & Kilian Q. Weinberger. * Learning from Simulated and Unsupervised Images through Adversarial Training [1] by Ashish Shrivastava, Tomas Pfister, Oncel Tuzel, Joshua Susskind, Wenda…

      iraphaelView on HNParent context

  6. Supercharge your Computer Vision models with the TensorFlow Object Detection API 1 comment

    • Their repository is pretty neat! It includes three state-of-the-art architectures in object detection: Faster-RCNN, RFCN, and SSD. It is missing YOLO [1][2], though, which shares some similarities with SSD. Another detector is the recently released Mask-RCNN [3], which of course…

      yamanekoView on HNParent context

Submitted to Hacker News

Connected through shared HN threads, not inferred similarity or a recommendation.