Dayoung Kil

Dayoung Kil

I am Dayoung, a Ph.D. student at the Graduate School of Soongsil University. My research primarily focuses on hardware-efficient artificial intelligence.

Publications

You can also find my articles on my Google Scholar profile.

Conference Papers

One-Cycle Structured Pruning via Stability-Driven Subnetwork Search

Published in The IEEE/CVF Winter Conference on Applications of Computer Vision 2026 (WACV), 2025

The key idea is to identify the optimal sub-network early in training, guided by norm-based group saliency criteria and structured sparsity regularization. We further introduce a novel pruning indicator that determines the stable pruning epoch by measuring the similarity between evolving sub-networks across consecutive epochs. The group sparsity regularization accelerates pruning, further reducing training time.

A Study of Structure Pruning for Hybrid Neural Networks

Published in 2024 24th International Conference on Control, Automation and Systems (ICCAS), 2024

In this paper, we explore the impact of structure pruning on model compression. Structured pruning, which removes specific structures within the model such as entire neurons, channels, or filters in convolutional neural networks, targets particular elements for removal.

단일 파노라마 입력의 실내 공간 레이아웃 복원 모델 경량화 [KCI]

Published in Journal of Institute of Control, Robotics and Systems (ICROS), 2022

In this paper, we present a lightweight deep learning model for room layout estimation. In contrast to the baseline method that uses a combination of a typical residual network (ResNet) and long short-term memory (LSTM) as its principal architecture, we focus on the use of a platform-aware neural architecture search for mobile applications (MnasNet) and a gated recurrent unit (GRU) instead of conventional LSTM.