Zobrazeno 1 - 10
of 446
pro vyhledávání: '"Lee, Jinkyu"'
In the training and inference of spiking neural networks (SNNs), direct training and lightweight computation methods have been orthogonally developed, aimed at reducing power consumption. However, only a limited number of approaches have applied thes
Externí odkaz:
http://arxiv.org/abs/2408.12293
Multi-object tracking (MOT) aims to construct moving trajectories for objects, and modern multi-object trackers mainly utilize the tracking-by-detection methodology. Initial approaches to MOT attacks primarily aimed to degrade the detection quality o
Externí odkaz:
http://arxiv.org/abs/2408.12727
Autor:
Lee, JinKyu, Kim, Jihie
Understanding commonsense knowledge is crucial in the field of Natural Language Processing (NLP). However, the presence of demographic terms in commonsense knowledge poses a potential risk of compromising the performance of NLP models. This study aim
Externí odkaz:
http://arxiv.org/abs/2406.07229
Streaming automatic speech recognition (ASR) models are restricted from accessing future context, which results in worse performance compared to the non-streaming models. To improve the performance of streaming ASR, knowledge distillation (KD) from t
Externí odkaz:
http://arxiv.org/abs/2308.16415
Autor:
Park, Jin Bok, Lee, Jinkyu, Back, Muhyun, Han, Hyunmin, Ma, David T., Won, Sang Min, Hwang, Sung Soo, Chun, Il Yong
In autonomous driving, the end-to-end (E2E) driving approach that predicts vehicle control signals directly from sensor data is rapidly gaining attention. To learn a safe E2E driving system, one needs an extensive amount of driving data and human int
Externí odkaz:
http://arxiv.org/abs/2308.14329
Autor:
Kang, Donghwa, Lee, Seunghoon, Chwa, Hoon Sung, Bae, Seung-Hwan, Kang, Chang Mook, Lee, Jinkyu, Baek, Hyeongboo
Different from existing MOT (Multi-Object Tracking) techniques that usually aim at improving tracking accuracy and average FPS, real-time systems such as autonomous vehicles necessitate new requirements of MOT under limited computing resources: (R1)
Externí odkaz:
http://arxiv.org/abs/2210.11946
A stagewise decomposition algorithm called value function gradient learning (VFGL) is proposed for large-scale multistage stochastic convex programs. VFGL finds the parameter values that best fit the gradient of the value function within a given para
Externí odkaz:
http://arxiv.org/abs/2205.08934
Autor:
Kim, Eunhyung, Lee, Jinkyu, Kim, Se-Jeong, Kim, Eun Mi, Byun, Hayeon, Huh, Seung Jae, Lee, Eunjin, Shin, Heungsoo
Publikováno v:
In Materials Today Bio December 2024 29
Autor:
Lee, Jaewoo, Lee, Jinkyu
Publikováno v:
In Future Generation Computer Systems November 2024 160:406-419
Autor:
Byun, Hayeon, Han, Yujin, Kim, Eunhyung, Jun, Indong, Lee, Jinkyu, Jeong, Hyewoo, Huh, Seung Jae, Joo, Jinmyoung, Shin, Su Ryon, Shin, Heungsoo
Publikováno v:
In Bioactive Materials June 2024 36:185-202