DNN Framework

Deep neural network (DNN) frameworks are the foundation of modern artificial intelligence, aiming to improve model accuracy, efficiency, and robustness. Current research focuses on optimizing DNN training and inference, including techniques like efficient parallelization strategies, mixed-precision training, and adaptive model selection based on resource constraints and carbon footprint. These advancements are crucial for deploying DNNs on resource-limited devices (e.g., edge computing) and mitigating challenges like adversarial attacks, noise, and data scarcity, ultimately impacting various fields from computer vision to natural language processing.

Papers

February 2, 2022

Auto-Transfer: Learning to Route Transferrable Representations
Keerthiram Murugesan, Vijay Sadashivaiah, Ronny Luss, Karthikeyan Shanmugam, Pin-Yu Chen, Amit Dhurandhar
Transfer Learning DNN Framework Transferable Representation Meta Representation Adversarial Multi Armed Bandit Target Representation

January 20, 2022

HEAM: High-Efficiency Approximate Multiplier Optimization for Deep Neural Networks
Su Zheng, Zhen Li, Yao Lu, Jingbo Gao, Jide Zhang, Lingli Wang
Deep Neural Network DNN Framework DNN Accelerator Class Wise Multiplier Accuracy Loss Approximate Multiplier

January 17, 2022

Black-box Error Diagnosis in Deep Neural Networks for Computer Vision: a Survey of Tools
Piero Fraternali, Federico Milani, Rocio Nahime Torres, Niccolò Zangrando
Deep Neural Network Timely Survey Computer Vision DNN Framework Easy Tool Error Classification

January 15, 2022

Explainability Tools Enabling Deep Learning in Future In-Situ Real-Time Planetary Explorations
Daniel Lundstrom, Alexander Huyen, Arya Mevada, Kyongsik Yun, Thomas Lu
Deep Learning Deep Neural Network DNN Framework Machine Learning Framework Planetary Exploration Integrated Gradient Planetary Rover

December 24, 2021

BMPQ: Bit-Gradient Sensitivity Driven Mixed-Precision Quantization of DNNs from Scratch
Souvik Kundu, Shikai Wang, Qirui Sun, Peter A. Beerel, Massoud Pedram
Deep Neural Network DNN Framework Scratch Project Mixed Precision Quantization Tiny ImageNet Gradient Coding Mixed Precision Training

December 19, 2021

Deep Filtering with DNN, CNN and RNN
Bin Xie, Qing Zhang
Deep Learning Approach DNN Framework CNN Network \Sigma}{\Delta}$ Low Pas RNN Deep Filter

December 18, 2021

LegoDNN: Block-grained Scaling of Deep Neural Networks for Mobile Vision
Rui Han, Qinglong Zhang, Chi Harold Liu, Guoren Wang, Jian Tang, Lydia Y. Chen
Deep Neural Network DNN Framework DNN Model Multiplicative Size Scaling Multiple Deep Neural Network DNN Workload Mobile Vision

November 21, 2021

Differentiable Projection for Constrained Deep Learning
Dou Huang, Haoran Zhang, Xuan Song, Ryosuke Shibasaki
Neural Network Deep Neural Network DNN Framework Differentiable Projection

November 11, 2021

November 8, 2021

Robust and Information-theoretically Safe Bias Classifier against Adversarial Attacks
Lijia Yu, Xiao-Shan Gao
Adversarial Attack Robust Version DNN Framework Gradient Based Attack Biased Classifier Safety Classifier

November 5, 2021

DNN Framework

Papers

Auto-Transfer: Learning to Route Transferrable Representations

HEAM: High-Efficiency Approximate Multiplier Optimization for Deep Neural Networks

Black-box Error Diagnosis in Deep Neural Networks for Computer Vision: a Survey of Tools

Explainability Tools Enabling Deep Learning in Future In-Situ Real-Time Planetary Explorations

BMPQ: Bit-Gradient Sensitivity Driven Mixed-Precision Quantization of DNNs from Scratch

Deep Filtering with DNN, CNN and RNN

LegoDNN: Block-grained Scaling of Deep Neural Networks for Mobile Vision

Differentiable Projection for Constrained Deep Learning

Variability-Aware Training and Self-Tuning of Highly Quantized DNNs for Analog PIM

Defining and Quantifying the Emergence of Sparse Concepts in DNNs

Robust and Information-theoretically Safe Bias Classifier against Adversarial Attacks

Interpreting Representation Quality of DNNs for 3D Point Cloud Processing

Visualizing the Emergence of Intermediate Visual Patterns in DNNs