Mimicking Neuro-inspired ArchitectureS in Hardware の履歴(No.8)

• 履歴一覧
• 差分 を表示
• 現在との差分 を表示
• ソース を表示
• Mimicking Neuro-inspired ArchitectureS in Hardware へ行く。
  • 1 (2017-02-21 (火) 03:35:10)
  • 2 (2017-02-21 (火) 05:44:05)
  • 3 (2017-02-21 (火) 09:19:19)
  • 4 (2017-02-24 (金) 07:50:23)
  • 5 (2017-03-09 (木) 06:05:17)
  • 6 (2017-03-09 (木) 10:25:02)
  • 7 (2017-03-10 (金) 07:59:24)
  • 8 (2017-03-10 (金) 08:02:55)

<== Back to NASH Project

Pig Face Recognition on FPGA using Deep Convolution Neural Network†

Leader: Murakami-1

Motivation†

Deep Neural Networks have been proved to be powerful tools for real world applications/tasks, such as pattern recognition, classification, regression, and prediction. However, simulating a large network in real-time requires high-performance machines or accelerators. Typical accelerators for large-scale NN accelerators use GPUs or ASIC chips. While ASICs deliver high performance, they lack the flexibility to reconfigure and hence are unable to adapt variation in the design and models employed. GPUs have better speedup over multi-core CPUs and good flexibility, but it lacks scalability to handle larger networks.

Goal and Expected output†

The goal of this research is to implement an FPGA-based Convolutional Neural Network for Pig Recognition.

• References
  • 2.Cat Face Recognition with CNN

• 5.Visualizing and Understanding Convolutional Networks

Implementation of a Recurrent Neural Network on FPGA for Video Description Generation or (Image Caption Generation)†

Looking at a sequence of images and find out what is going on.

• References

• 1.Introduction to DL https://www.youtube.com/watch?v=1L0TKZQcUtA

• 2.Multi-layer Recurrent Neural Networks (LSTM, GRU, RNN) for character-level language models in Torch https://github.com/karpathy/char-rnn

• 3.Sequence to Sequence – Video to Text

Modeling attention Steering Application: Example: When a drone flies at a very high speed (+300 fps), it cannot capture the whole scene. So, what we can do is steering around an image)

• Multiple Object Recognition with Visual Attention, 2015
• DRAW: A Recurrent Neural Network For Image Generation

DL Course†

• http://selfdrivingcars.mit.edu/
• http://cs231n.github.io/