About Me

Edwin Arkel Rios
PhD Student in EECS
National Yang Ming Chiao Tung University
Hsinchu, Taiwan

edwinarkel.rios@gmail.com
[GitHub] [Resume]

My name's Edwin and I come from Panama. I received the B.S. degree in Energy Engineering from National Cheng Kung University, Tainan, Taiwan, in 2019. I enrolled National Chiao Tung University to pursue a M.S. and then a Ph.D. degree in Electrical Engineering and Computer Science, in 2019 and 2021, respectively, where I have been co-supervised by Professor Bo-Cheng Lai and Professor Min-Chun Hu.

My research is focused on the design of efficient deep learning models. In particular, I have done work on time-series for heart-rate monitoring, model compression using knowledge distillation, and fine-grained image recognition.

Publications

Global-Local Similarity for Efficient Fine-Grained Image Recognition with Vision Transformers
Similarity metric between global and local representation as criteria for discriminative region selection criteria for FGIR.
Thorough experiments across 14 FGIR datasets covering variety of settings including image size and AugReg combinations.
Our proposed model obtains best accuracy on 8 / 14 at much less cost than best on others.
[Source code] [Paper (coming soon)]
Anime Character Recognition using Intermediate Features Aggregation. ISCAS 2022
Intermediate features aggregation classification head for reducing sensitivity to hyperparameters in ViTs at low computational cost.
Experiments on how model architecture and hyperparameter affect classification performance.
[Source code] [Paper]
IFACD: Intermediate Features Augmented Contrastive Distillation. ICLR CSS Workshop 2022
Extend contrastive learning framework (in particular SimCLR-like) to support features from intermediate layers of a particular image as additional positive pairs and intermediate features from other images in batch as negative pairs, reducing reliance on augmentations pairs and large batch size requirements.
Apply proposed contrastive loss in a knowledge distillation framework where intermediate features are extracted using a frozen teacher and student learns using a combined contrastive, distillation, and supervised cross-entropy loss.
Experiments with a variety of teacher-student combinations show our method obtains favorable accuracy at a competitive computational cost compared to alternatives.
[Source Code] [Recorded Presentation (Starts at 2:47:50)] [Slides]
DLPrPPG: Development and Design of Deep Learning Platform for Remote Photoplethysmography. ISCAS 2022
Parametric Study of Performance of Remote Photopletysmography System. ISCAS 2021
Systematic studies of remote photoplethysmography (RPPG) design hyperparameters (face detection method, region of interest selection scheme, sliding window size, preprocessing and postprocessing algorithm choice) and their effects on heart rate prediction accuracy in relation to computational cost using C++, Python and OpenCV.
[Paper ISCAS 2022] [Paper ISCAS 2021]

Projects

Vision Transformers for Anime Character Face Recognition
Image classification of anime character faces using state-of-the-art for image recognition: vision transformers.
Revamped a dataset for this task, DAF:re: almost 500 K images and more than 3 K classes.
Conducted experiments on both CNN and self-attention based models, and explored for the first-time the hyperparameter tuning (mini-batch and image size) of vision transformers for out-of-domain data.
[Source code] [Paper on DAF:re]
DeblurGAN for Deblurring of Videos
Done in Python, using OpenCV and PyTorch.
Final project for Digital Image Processing course (108-1).
Also included slides for midterm presentation covering DeblurGANv2 paper.
[Final Project Poster] [DeblurGANv2 Paper Slides]
Computer Vision for Autonomous Driving
Done in Python, making use of TensorFlow, Keras and OpenCV, and ROS.
Final project for Deep Learning course (108-1).
Also included slides for midterm presentation covering VGG network paper.
[VGG Paper Slides] [Final Project Slides]