Hardware Architectures for Deep Learning

Herausgeber: Daneshtalab, Masoud; Modarressi, Mehdi

• Gebundenes Buch

Produktbeschreibung

This book discusses innovative ideas in the design, modelling, implementation, and optimization of hardware platforms for neural networks. The book provides an overview of this emerging field, from principles to applications, for researchers, postgraduate students and engineers who work on learning-based services and hardware platforms.

Produktdetails

• Materials, Circuits and Device
• Verlag: Institution of Engineering & Technology
• Seitenzahl: 328
• Erscheinungstermin: 24. April 2020
• Englisch
• Abmessung: 239mm x 160mm x 20mm
• Gewicht: 680g
• ISBN-13: 9781785617683
• ISBN-10: 1785617680
• Artikelnr.: 57096683

Inhaltsangabe

1. Part I: Deep learning and neural networks: concepts and models
* Chapter 1: An introduction to artificial neural networks
* Chapter 2: Hardware acceleration for recurrent neural networks
* Chapter 3: Feedforward neural networks on massively parallel
architectures
2. Part II: Deep learning and approximate data representation
* Chapter 4: Stochastic-binary convolutional neural networks with
deterministic bit-streams
* Chapter 5: Binary neural networks
3. Part III: Deep learning and model sparsity
* Chapter 6: Hardware and software techniques for sparse deep neural
networks
* Chapter 7: Computation reuse-aware accelerator for neural networks
4. Part IV: Convolutional neural networks for embedded systems
* Chapter 8: CNN agnostic accelerator design for low latency inference
on FPGAs
* Chapter 9: Iterative convolutional neural network (ICNN): an
iterative CNN solution for low power and real-time systems
5. Part V: Deep learning on analog accelerators
* Chapter 10: Mixed-signal neuromorphic platform design for streaming
biomedical signal processing
* Chapter 11: Inverter-based memristive neuromorphic circuit for
ultra-low-power IoT smart applications