Presentation is loading. Please wait.

Presentation is loading. Please wait.

Faster R-CNN – Concepts

Similar presentations


Presentation on theme: "Faster R-CNN – Concepts"— Presentation transcript:

1 Faster R-CNN – Concepts
Student Presentation by: Assaf Livne Based on the work of: Ross Girdhick, Shaoqing Ren, Kaiming He אוניברסיטת בן-גוריון בנגב Ben-Gurion University of the Negev Faculty of Engineering Sciences Department of Electrical Engineering

2 Introduction R-CNN Concepts Fast R-CNN Concepts Faster R-CNN Concepts
5/12/2016 Introduction R-CNN Concepts Fast R-CNN Concepts Faster R-CNN Concepts Conclusion

3 17/11/2016 Introduction

4 Introduction ImageNet 17/11/2016

5 17/11/2016 Introduction Kitti

6 17/11/2016 R-CNN

7 17/11/2016 R-CNN concepts Lets combine Localization NN and classification NN in the simplest way.

8 17/11/2016

9 17/11/2016

10 Training Process Take a pre-trained classification network.
17/11/2016 Training Process Take a pre-trained classification network. Re-train the last fully connected layer with the objects that need to be detected + "no-object" class. Get all proposals(=~2000 p/image), resize them to match the cnn input, then save to disk. Train SVM to classify between object and background. BB Regression: Train a linear regression classifier that will output some correction factor.

11 17/11/2016 R-CNN drawbacks Numerous Candidate object locations must be processed – slow training time and test time Ad hoc training objective. correlation between image space location and detected class is developing.

12 17/11/2016 Fast R-CNN

13 17/11/2016 Fast R-CNN concepts Instead of running the CNN on every proposal lets try to save some resources. Train all the layers in a single stage. No memory consumption. Inspired from the VGG16 concept - Very Deep CNN.

14 17/11/2016 33

15 17/11/2016

16 17/11/2016 ROI pooling Type of max-pooling with a pool size dependent on the input, so that the output always has the same size. This is done because fully connected layer always expected the same input size.

17 17/11/2016

18 17/11/2016

19 17/11/2016

20 R-CNN vs Fast R-CNN Testing time: mAP (VOC 2007): 49s 2.32s 66% 66.9%
17/11/2016 R-CNN vs Fast R-CNN Testing time: 49s s mAP (VOC 2007): 66% %

21 17/11/2016 Fast R-CNN drawbacks Still depends on an external object proposal system . Which is the major bottleneck from computing resources point of view.

22 17/11/2016 Faster R-CNN

23 17/11/2016 Faster R-CNN concepts Using the already running CNN to infer region proposals.

24 Faster R-CNN Pipe-Lines
17/11/2016 Faster R-CNN Pipe-Lines Get feature maps from the deep convolution layers. Train a Region Proposal Network (RPN). Give proposals to the ROI pooling layer. Send proposals to a fully connected layer to finish the classification.

25 17/11/2016

26 Region proposal Network (RPN)
17/11/2016 Region proposal Network (RPN) Basically the RPN is a sliding window which slides on the feature map. Sends as an output the locations of the proposals windows.

27 Fast R-CNN vs Faster R-CNN
17/11/2016 Fast R-CNN vs Faster R-CNN Testing time: 2.32s s mAP (VOC 2007): 66.9% %

28 17/11/2016 Conclusion

29 17/11/2016 Conclusion “Using the recently popular terminology of neural networks with ’attention’ mechanisms, the RPN module tells the Fast R-CNN module where to look.”

30 ’Attention’ Mechanisms
17/11/2016 ’Attention’ Mechanisms Rather than using all available information, we need to select the most pertinent piece of information.

31 RNN – Encoder Decoder Model
17/11/2016 RNN – Encoder Decoder Model

32 RNN - Attention Model 17/11/2016

33 Natural Language Processing (NLP)
17/11/2016 Natural Language Processing (NLP) Show, Attend and Tell – Kelvin Xu et al 2015

34 Natural Language Processing (NLP)
17/11/2016 Natural Language Processing (NLP) Show, Attend and Tell – Kelvin Xu et al 2015

35 Thank you for your attention!
17/11/2016 Thank you for your attention!


Download ppt "Faster R-CNN – Concepts"

Similar presentations


Ads by Google