Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Neural Regulation01:37

Neural Regulation

40.6K
Digestion begins with a cephalic phase that prepares the digestive system to receive food. When our brain processes visual or olfactory information about food, it triggers impulses in the cranial nerves innervating the salivary glands and stomach to prepare for food.
40.6K
Neural Circuits01:25

Neural Circuits

2.0K
Neural circuits and neuronal pools are two of the main structures found in the nervous system. Neural circuits are networks of neurons that work together to carry out a specific task or process. They consist of interconnected neurons and glial cells, which provide structural and metabolic support.
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
2.0K

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Chameleon: Adaptive Code Optimization for Expedited Deep Neural Network Compilation.

... International Conference on Learning Representations·2026
Same authorSame journal

Yin-Yang: Programming Abstractions for Cross-Domain Multi-Acceleration.

IEEE micro·2023
Same author

High Density Individually Addressable Nanowire Arrays Record Intracellular Activity from Primary Rodent and Human Stem Cell Derived Neurons.

Nano letters·2017
See all related articles

Related Experiment Video

Updated: Oct 23, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.5K

ReLeQ : A Reinforcement Learning Approach for Automatic Deep Quantization of Neural Networks.

Ahmed T Elthakeb1, Prannoy Pilligundla1, Fatemehsadat Mireshghallah1

  • 1University of California San Diego.

IEEE Micro
|August 20, 2021
PubMed
Summary
This summary is machine-generated.

Deep Quantization below eight bits reduces DNN computation and storage. ReLeQ, a reinforcement learning framework, automates bitwidth discovery, minimizing accuracy loss and maximizing speedup.

More Related Videos

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
09:47

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

1.4K
A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis
05:41

A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis

Published on: February 6, 2020

9.6K

Related Experiment Videos

Last Updated: Oct 23, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.5K
Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
09:47

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

1.4K
A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis
05:41

A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis

Published on: February 6, 2020

9.6K

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Deep Quantization (below eight bits) significantly reduces Deep Neural Network (DNN) computation and storage.
  • Manual deep quantization often results in substantial accuracy loss, limiting its practical application.

Purpose of the Study:

  • To systematically automate the discovery of optimal bitwidths for deep quantization.
  • To minimize accuracy loss and computational/storage costs in quantized DNNs.

Main Methods:

  • Developed an end-to-end deep reinforcement learning framework named ReLeQ.
  • Utilized proximal policy optimization for efficient exploration of bitwidth assignments.
  • Applied heterogeneous bitwidth assignment to various deep networks.

Main Results:

  • Achieved minimal accuracy loss (≤ 0.3%) across diverse deep networks.
  • Demonstrated significant reduction in computation and storage costs.
  • Enabled a 2.2× speedup over 8-bit execution on conventional and custom hardware.

Conclusions:

  • ReLeQ offers a systematic and automated approach to deep quantization.
  • The framework effectively balances network speed and quality.
  • ReLeQ enhances the utility of deep quantization for efficient DNN deployment.