activity
20172020
most citedBinaryDuo: Reducing Gradient Mismatch in Binary Activation Network by Coupling Binary Activations

31 citations · 36 across the 4 of their papers we have counts for

collaborators

6 papers

cs.LG20205 cited

Improving Accuracy of Binary Neural Networks using Unbalanced Activation Distribution

Hyungjun Kim, Jihoon Park, Changhun Lee +1

Binarization of neural network models is considered as one of the promising methods to deploy deep neural network models on resource-constrained environments such as mobile devices…

cs.LG202031 cited

BinaryDuo: Reducing Gradient Mismatch in Binary Activation Network by Coupling Binary Activations

Hyungjun Kim, Kyungsu Kim, Jinseok Kim +1

Binary Neural Networks (BNNs) have been garnering interest thanks to their compute cost reduction and memory savings. However, BNNs suffer from performance degradation mainly due t…

cs.ET2019

Zero-shifting Technique for Deep Neural Network Training on Resistive Cross-point Arrays

Hyungjun Kim, Malte Rasch, Tayfun Gokmen +5

A resistive memory device-based computing architecture is one of the promising platforms for energy-efficient Deep Neural Network (DNN) training accelerators. The key technical cha…

cs.NE2019

BitSplit-Net: Multi-bit Deep Neural Network with Bitwise Activation Function

Hyungjun Kim, Yulhwa Kim, Sungju Ryu +1

Significant computational cost and memory requirements for deep neural networks (DNNs) make it difficult to utilize DNNs in resource-constrained environments. Binary neural network…

cs.NE2018

Neural Network-Hardware Co-design for Scalable RRAM-based BNN Accelerators

Yulhwa Kim, Hyungjun Kim, Jae-Joon Kim

Recently, RRAM-based Binary Neural Network (BNN) hardware has been gaining interests as it requires 1-bit sense-amp only and eliminates the need for high-resolution ADC and DAC. Ho…

cs.ET2017

Deep Neural Network Optimized to Resistive Memory with Nonlinear Current-Voltage Characteristics

Hyungjun Kim, Taesu Kim, Jinseok Kim +1

Artificial Neural Network computation relies on intensive vector-matrix multiplications. Recently, the emerging nonvolatile memory (NVM) crossbar array showed a feasibility of impl…