Picture for Mahdi Nazemi

Mahdi Nazemi

Memory-Efficient Vision Transformers: An Activation-Aware Mixed-Rank Compression Strategy

Add code
Feb 08, 2024
Viaarxiv icon

Low-Precision Mixed-Computation Models for Inference on Edge

Add code
Dec 03, 2023
Figure 1 for Low-Precision Mixed-Computation Models for Inference on Edge
Figure 2 for Low-Precision Mixed-Computation Models for Inference on Edge
Figure 3 for Low-Precision Mixed-Computation Models for Inference on Edge
Figure 4 for Low-Precision Mixed-Computation Models for Inference on Edge
Viaarxiv icon

Sensitivity-Aware Mixed-Precision Quantization and Width Optimization of Deep Neural Networks Through Cluster-Based Tree-Structured Parzen Estimation

Add code
Aug 16, 2023
Viaarxiv icon

SNT: Sharpness-Minimizing Network Transformation for Fast Compression-friendly Pretraining

Add code
May 08, 2023
Figure 1 for SNT: Sharpness-Minimizing Network Transformation for Fast Compression-friendly Pretraining
Figure 2 for SNT: Sharpness-Minimizing Network Transformation for Fast Compression-friendly Pretraining
Figure 3 for SNT: Sharpness-Minimizing Network Transformation for Fast Compression-friendly Pretraining
Figure 4 for SNT: Sharpness-Minimizing Network Transformation for Fast Compression-friendly Pretraining
Viaarxiv icon

A Fast Training-Free Compression Framework for Vision Transformers

Add code
Mar 04, 2023
Figure 1 for A Fast Training-Free Compression Framework for Vision Transformers
Figure 2 for A Fast Training-Free Compression Framework for Vision Transformers
Figure 3 for A Fast Training-Free Compression Framework for Vision Transformers
Figure 4 for A Fast Training-Free Compression Framework for Vision Transformers
Viaarxiv icon

Efficient Compilation and Mapping of Fixed Function Combinational Logic onto Digital Signal Processors Targeting Neural Network Inference and Utilizing High-level Synthesis

Add code
Jul 30, 2022
Figure 1 for Efficient Compilation and Mapping of Fixed Function Combinational Logic onto Digital Signal Processors Targeting Neural Network Inference and Utilizing High-level Synthesis
Figure 2 for Efficient Compilation and Mapping of Fixed Function Combinational Logic onto Digital Signal Processors Targeting Neural Network Inference and Utilizing High-level Synthesis
Figure 3 for Efficient Compilation and Mapping of Fixed Function Combinational Logic onto Digital Signal Processors Targeting Neural Network Inference and Utilizing High-level Synthesis
Figure 4 for Efficient Compilation and Mapping of Fixed Function Combinational Logic onto Digital Signal Processors Targeting Neural Network Inference and Utilizing High-level Synthesis
Viaarxiv icon

NullaNet Tiny: Ultra-low-latency DNN Inference Through Fixed-function Combinational Logic

Add code
Apr 07, 2021
Figure 1 for NullaNet Tiny: Ultra-low-latency DNN Inference Through Fixed-function Combinational Logic
Figure 2 for NullaNet Tiny: Ultra-low-latency DNN Inference Through Fixed-function Combinational Logic
Viaarxiv icon

A Tunable Robust Pruning Framework Through Dynamic Network Rewiring of DNNs

Add code
Nov 03, 2020
Figure 1 for A Tunable Robust Pruning Framework Through Dynamic Network Rewiring of DNNs
Figure 2 for A Tunable Robust Pruning Framework Through Dynamic Network Rewiring of DNNs
Figure 3 for A Tunable Robust Pruning Framework Through Dynamic Network Rewiring of DNNs
Figure 4 for A Tunable Robust Pruning Framework Through Dynamic Network Rewiring of DNNs
Viaarxiv icon

SynergicLearning: Neural Network-Based Feature Extraction for Highly-Accurate Hyperdimensional Learning

Add code
Aug 04, 2020
Figure 1 for SynergicLearning: Neural Network-Based Feature Extraction for Highly-Accurate Hyperdimensional Learning
Figure 2 for SynergicLearning: Neural Network-Based Feature Extraction for Highly-Accurate Hyperdimensional Learning
Figure 3 for SynergicLearning: Neural Network-Based Feature Extraction for Highly-Accurate Hyperdimensional Learning
Figure 4 for SynergicLearning: Neural Network-Based Feature Extraction for Highly-Accurate Hyperdimensional Learning
Viaarxiv icon

Pre-defined Sparsity for Low-Complexity Convolutional Neural Networks

Add code
Feb 04, 2020
Figure 1 for Pre-defined Sparsity for Low-Complexity Convolutional Neural Networks
Figure 2 for Pre-defined Sparsity for Low-Complexity Convolutional Neural Networks
Figure 3 for Pre-defined Sparsity for Low-Complexity Convolutional Neural Networks
Figure 4 for Pre-defined Sparsity for Low-Complexity Convolutional Neural Networks
Viaarxiv icon