Seunghyun Lee 👋 
- Welcome to my Github page. I am a Ph.D. course student at Inha Univ. in South Korea. My research areas are machine learning, deep learning, and especially the light-weighting the convolutional neural networks such as knowledge distillation and filter pruning.
ML libraries 🧱
- Tensorflow (1.x and 2.x): Upper intermediate
- Pytorch: intermediate
- JAX: beginner
Academic activity 🕹
-
Leader of deep learning paper study group: link
-
Major contributor of the implementation project for Putting NeRF on a Diet
in
🤗 HuggingFace X GoogleAI Flax/JAX Community Week Event (won the 2nd price!😆 ) -
Have served as a reviewer for Neural Networks.
Publication 📜
First author of
- "Self-supervised Knowledge Distillation Using Singular Value Decomposition" on ECCV2018 [paper] [TF1 code, TF2 code]
- "Graph-based Knowledge Distillation by Multi-head Attention Network." on BMVC2019 oral [paper] [code]
- "Transformation of Non-Euclidean Space to Euclidean Space for Efficient Learning of Singular Vectors" on IEEE Access (2020) [paper]
- "Filter Pruning and Re-Initialization via Latent Space Clustering" on IEEE Access (2020) [paper]
- "Knowledge Transfer via Decomposing Essential Information in Convolutional Neural Networks" on IEEE TNNLS (2020) [paper] [TF1 code, TF2 code]
- "Interpretable Embedding Procedure Knowledge Transfer via Stacked Principal Component Analysis and Graph Neural Network" on AAAI2021 [paper] [code]
- "Ensemble Knowledge Guided Sub-network Search and Fine-tuning for Filter Pruning" on arxiv preprint [paper] [code]
Co-author of
- "MUNet: macro unit-based convolutional neural network for mobile devices" on CVPRW2018 [paper]
- "Metric-Based Regularization and Temporal Ensemble for Multi-Task Learning using Heterogeneous Unsupervised Tasks" on ICCVW2019 [paper]
- "Real-time purchase behavior recognition system based on deep learning-based object detection and tracking for an unmanned product cabinet" on ESWA (2020) [paper]
- "Channel Pruning Via Gradient Of Mutual Information For Light-Weight Convolutional Neural Networks" on ICIP 2020 [paper]
- "Zero-Shot Knowledge Distillation Using Label-Free Adversarial Perturbation With Taylor Approximation" on IEEE Access (2021) [paper] [code]
- "Contextual Gradient Scaling for Few-Shot Learning" on WACV2022 [paper] [code]
- "Vision Transformer for Small-Size Datasets" on arxiv preprint [paper] [code]