Deep Learning for Computer Vision: Object Detection (UPC 2016)
11 likes6,501 views
The document discusses advancements in object detection techniques, focusing on deep convolutional networks and proposals for region detection. It compares various models including R-CNN, Fast R-CNN, and Faster R-CNN, highlighting their speed and accuracy improvements over time. Additionally, it touches on architectures like YOLO and SSD, and presents training resources related to these methods.
![[course site]
Object Detection
Day 3 Lecture 4
Amaia Salvador
amaia.salvador@upc.edu](https://image.slidesharecdn.com/dlcvd3l4objects-160803161336/85/Deep-Learning-for-Computer-Vision-Object-Detection-UPC-2016-1-320.jpg)
![Object Detection as Classification
Classes = [cat, dog, duck]
Cat ? NO
Dog ? NO
Duck? NO
4](https://image.slidesharecdn.com/dlcvd3l4objects-160803161336/85/Deep-Learning-for-Computer-Vision-Object-Detection-UPC-2016-4-320.jpg)
![Object Detection as Classification
Classes = [cat, dog, duck]
Cat ? NO
Dog ? NO
Duck? NO
5](https://image.slidesharecdn.com/dlcvd3l4objects-160803161336/85/Deep-Learning-for-Computer-Vision-Object-Detection-UPC-2016-5-320.jpg)
![Object Detection as Classification
Classes = [cat, dog, duck]
Cat ? YES
Dog ? NO
Duck? NO
6](https://image.slidesharecdn.com/dlcvd3l4objects-160803161336/85/Deep-Learning-for-Computer-Vision-Object-Detection-UPC-2016-6-320.jpg)
![Object Detection as Classification
Classes = [cat, dog, duck]
Cat ? NO
Dog ? NO
Duck? NO
7](https://image.slidesharecdn.com/dlcvd3l4objects-160803161336/85/Deep-Learning-for-Computer-Vision-Object-Detection-UPC-2016-7-320.jpg)
![Region Proposals
Selective Search (SS) Multiscale Combinatorial Grouping (MCG)
[SS] Uijlings et al. Selective search for object recognition. IJCV 2013
[MCG] Arbeláez, Pont-Tuset et al. Multiscale combinatorial grouping. CVPR 2014 13](https://image.slidesharecdn.com/dlcvd3l4objects-160803161336/85/Deep-Learning-for-Computer-Vision-Object-Detection-UPC-2016-13-320.jpg)
![Resources
● Related Lecture from CS231n @ Stanford [slides][video]
● Caffe Code for:
○ R-CNN
○ Fast R-CNN
○ Faster R-CNN [matlab][python]
● YOLO
○ Original (Darknet)
○ Tensorflow
○ Keras
● SSD (Caffe)
31](https://image.slidesharecdn.com/dlcvd3l4objects-160803161336/85/Deep-Learning-for-Computer-Vision-Object-Detection-UPC-2016-31-320.jpg)
Deep Learning for Computer Vision: Object Detection (UPC 2016)
- 1. [course site] Object Detection Day 3 Lecture 4 Amaia Salvador [email protected]
- 2. Slide Credit: Xavier Giró Images (global) Objects (local) Deep ConvNets for Recognition for... Video (2D+T) 2
- 3. Object Detection CAT, DOG, DUCK The task of assigning a label and a bounding box to all objects in the image 3
- 4. Object Detection as Classification Classes = [cat, dog, duck] Cat ? NO Dog ? NO Duck? NO 4
- 5. Object Detection as Classification Classes = [cat, dog, duck] Cat ? NO Dog ? NO Duck? NO 5
- 6. Object Detection as Classification Classes = [cat, dog, duck] Cat ? YES Dog ? NO Duck? NO 6
- 7. Object Detection as Classification Classes = [cat, dog, duck] Cat ? NO Dog ? NO Duck? NO 7
- 8. Object Detection as Classification Problem: Too many positions & scales to test Solution: If your classifier is fast enough, go for it 8
- 9. HOG Dalal and Triggs. Histograms of Oriented Gradients for Human Detection. CVPR 2005 9
- 10. Deformable Part Model Felzenszwalb et al, Object Detection with Discriminatively Trained Part Based Models, PAMI 2010 10
- 11. Object Detection with CNNs? CNN classifiers are computationally demanding. We can’t test all positions & scales ! Solution: Look at a tiny subset of positions. Choose them wisely :) 11
- 12. Region Proposals ● Find “blobby” image regions that are likely to contain objects ● “Class-agnostic” object detector ● Look for “blob-like” regions Slide Credit: CS231n 12
- 13. Region Proposals Selective Search (SS) Multiscale Combinatorial Grouping (MCG) [SS] Uijlings et al. Selective search for object recognition. IJCV 2013 [MCG] Arbeláez, Pont-Tuset et al. Multiscale combinatorial grouping. CVPR 2014 13
- 14. Object Detection with CNNs: R-CNN Girshick et al. Rich feature hierarchies for accurate object detection and semantic segmentation. CVPR 2014 14
- 15. R-CNN Girshick et al. Rich feature hierarchies for accurate object detection and semantic segmentation. CVPR 2014 1. Train network on proposals 2. Post-hoc training of SVMs & Box regressors on fc7 features 15
- 16. R-CNN Girshick et al. Rich feature hierarchies for accurate object detection and semantic segmentation. CVPR 2014 16
- 17. R-CNN: Problems 1. Slow at test-time: need to run full forward pass of CNN for each region proposal 2. SVMs and regressors are post-hoc: CNN features not updated in response to SVMs and regressors 3. Complex multistage training pipeline Slide Credit: CS231n 17
- 18. Fast R-CNN Girshick Fast R-CNN. ICCV 2015 Solution: Share computation of convolutional layers between region proposals for an image R-CNN Problem #1: Slow at test-time: need to run full forward pass of CNN for each region proposal 18
- 19. Fast R-CNN Hi-res input image: 3 x 800 x 600 with region proposal Convolution and Pooling Hi-res conv features: C x H x W with region proposal Fully-connected layers Max-pool within each grid cell RoI conv features: C x h x w for region proposal Fully-connected layers expect low-res conv features: C x h x w Slide Credit: CS231n 19Girshick Fast R-CNN. ICCV 2015
- 20. Fast R-CNN Solution: Train it all at together E2E R-CNN Problem #2&3: SVMs and regressors are post-hoc. Complex training. 20Girshick Fast R-CNN. ICCV 2015
- 21. Fast R-CNN Slide Credit: CS231n R-CNN Fast R-CNN Training Time: 84 hours 9.5 hours (Speedup) 1x 8.8x Test time per image 47 seconds 0.32 seconds (Speedup) 1x 146x mAP (VOC 2007) 66.0 66.9 Using VGG-16 CNN on Pascal VOC 2007 dataset Faster! FASTER! Better! 21
- 22. Fast R-CNN: Problem Slide Credit: CS231n R-CNN Fast R-CNN Test time per image 47 seconds 0.32 seconds (Speedup) 1x 146x Test time per image with Selective Search 50 seconds 2 seconds (Speedup) 1x 25x Test-time speeds don’t include region proposals 22
- 23. Faster R-CNN Conv layers Region Proposal Network FC6 Class probabilities FC7 FC8 RPN Proposals RoI Pooling Conv5_3 RPN Proposals 23 Ren et al. Faster R-CNN: Towards real-time object detection with region proposal networks. NIPS 2015
- 24. Faster R-CNN Conv layers Region Proposal Network FC6 Class probabilities FC7 FC8 RPN Proposals RoI Pooling Conv5_3 RPN Proposals Fast R-CNN 24 Ren et al. Faster R-CNN: Towards real-time object detection with region proposal networks. NIPS 2015
- 25. Region Proposal Network Objectness scores (object/no object) Bounding Box Regression In practice, k = 9 (3 different scales and 3 aspect ratios) 25 Ren et al. Faster R-CNN: Towards real-time object detection with region proposal networks. NIPS 2015
- 26. Faster R-CNN Ren et al. Faster R-CNN: Towards real-time object detection with region proposal networks. NIPS 2015 R-CNN Fast R-CNN Faster R-CNN Test time per image (with proposals) 50 seconds 2 seconds 0.2 seconds (Speedup) 1x 25x 250x mAP (VOC 2007) 66.0 66.9 66.9 Slide Credit: CS231n 26
- 27. Faster R-CNN 27 ● Faster R-CNN is the basis of the winners of COCO and ILSVRC 2015 object detection competitions. He et al. Deep residual learning for image recognition. arXiv 2015
- 28. YOLO: You Only Look Once Slide Credit: CS231n Divide image into S x S grid Within each grid cell predict: B Boxes: 4 coordinates + confidence Class scores: C numbers Regression from image to 7 x 7 x (5 * B + C) tensor Direct prediction using a CNN Redmon et al. You Only Look Once: Unified, Real-Time Object Detection, CVPR 2016 28
- 29. SSD: Single Shot MultiBox Detector Liu et al. SSD: Single Shot MultiBox Detector, arXiv 2015 29
- 30. SSD: Single Shot MultiBox Detector Liu et al. SSD: Single Shot MultiBox Detector, arXiv 2015 System VOC2007 test mAP FPS (Titan X) Number of Boxes Faster R-CNN (VGG16) 73.2 7 300 Faster R-CNN (ZF) 62.1 17 300 YOLO 63.4 45 98 Fast YOLO 52.7 155 98 SSD300 (VGG) 72.1 58 7308 SSD300 (VGG, cuDNN v5) 72.1 72 7308 SSD500 (VGG16) 75.1 23 20097 30 Training with Pascal VOC 07+12
- 31. Resources ● Related Lecture from CS231n @ Stanford [slides][video] ● Caffe Code for: ○ R-CNN ○ Fast R-CNN ○ Faster R-CNN [matlab][python] ● YOLO ○ Original (Darknet) ○ Tensorflow ○ Keras ● SSD (Caffe) 31