activity
20182021
most citedQuantization-Guided Training for Compact TinyML Models

5 citations · 5 across the 4 of their papers we have counts for

collaborators

6 papers

cs.LG20215 cited

Quantization-Guided Training for Compact TinyML Models

Sedigh Ghamari, Koray Ozcan, Thu Dinh +4

We propose a Quantization Guided Training (QGT) method to guide DNN training towards optimized low-bit-precision targets and reach extreme compression levels below 8-bit precision.…

cs.CV2020

Subtensor Quantization for Mobilenets

Thu Dinh, Andrey Melnikov, Vasilios Daskalopoulos +1

Quantization for deep neural networks (DNN) have enabled developers to deploy models with less memory and more efficient low-power inference. However, not all DNN designs are frien…

cs.LG2020

Sparsity Meets Robustness: Channel Pruning for the Feynman-Kac Formalism Principled Robust Deep Neural Nets

Thu Dinh, Bao Wang, Andrea L. Bertozzi +1

Deep neural nets (DNNs) compression is crucial for adaptation to mobile devices. Though many successful algorithms exist to compress naturally trained DNNs, developing efficient an…

math.OC2019

Convergence of a Relaxed Variable Splitting Coarse Gradient Descent Method for Learning Sparse Weight Binarized Activation Neural Networks

Thu Dinh, Jack Xin

Sparsification of neural networks is one of the effective complexity reduction methods to improve efficiency and generalizability. Binarized activation offers an additional computa…

math.OC2018

Convergence of a Relaxed Variable Splitting Method for Learning Sparse Neural Networks via , and transformed- Penalties

Thu Dinh, Jack Xin

Sparsification of neural networks is one of the effective complexity reduction methods to improve efficiency and generalizability. We consider the problem of learning a one hidden…

math.PR2018

Enhanced Diffusivity in Perturbed Senile Reinforced Random Walk Models

Thu Dinh, Jack Xin

We consider diffusivity of random walks with transition probabilities depending on the number of consecutive traversals of the last traversed edge, the so called senile reinforced…