Jinyoung Choi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jinyoung Choi has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
- Injung Kim
Handong Global University
9 shared publications - Dougho Park
Medical Science and Engineering, Graduate School of Convergence Science and Technology, Pohang University of Science and Technology, Pohang, South Korea
3 shared publications - Kihyun Na
Handong Global University
3 shared publications - Seokwon Kim
School of CSEE, Handong Global University, 558 Handong-ro, Heunghae-eup, Buk-gu, Pohang, Gyeongbuk 37554, Republic of Korea
2 shared publications - Seungeon Song
School of CSEE, Handong Global University, 558 Handong-ro, Heunghae-eup, Buk-gu, Pohang, Gyeongbuk 37554, Republic of Korea
2 shared publications - Taeyeon Kim
Medical Research Institute, Pohang Stroke and Spine Hospital, Pohang, South Korea
2 shared publications
Latest Publications
- 1 November 2026
- Published by Radiological Society of North America (RSNA) in Radiology: Artificial Intelligence
- Vol. 8 (6)
ARNAI: Artifact Removal Network Based on Autoencoding and Inpainting for Robust Spinal Image Segmentation and Measurement
A Restore, Segment, and Measure framework mitigated implant-related artifacts without implant-containing training data and improved automated spinopelvic assessment on postoperative lumbar radiographs.
- 1 November 2026
- Published by Radiological Society of North America (RSNA) in Radiology: Artificial Intelligence
- Vol. 8 (6)
ARNAI: Artifact Removal Network Based on Autoencoding and Inpainting for Robust Spinal Image Segmentation and Measurement
A Restore, Segment, and Measure framework mitigated implant-related artifacts without implant-containing training data and improved automated spinopelvic assessment on postoperative lumbar radiographs.
- 3 October 2026
- Published by Springer Nature
- p. 520-538
Rethinking Garment Conditioning in Diffusion-Based Virtual Try-On: Decouple, Don’t Denoise
- 3 October 2026
- Published by Springer Nature
- p. 520-538
Rethinking Garment Conditioning in Diffusion-Based Virtual Try-On: Decouple, Don’t Denoise
- 28 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 34, 2649-2662
BERT-APC: A Reference-Free Framework for Automatic Pitch Correction via Musical Context Inference
Automatic Pitch Correction (APC) enhances vocal recordings by correcting pitch deviations to align with the intended musical notes. However, existing APC systems either rely on reference pitches, which limits their practical applicability, or employ simple pitch estimation algorithms that often fail to preserve expressiveness and naturalness. We propose BERT-APC, a novel reference-free APC framework that corrects pitch errors while maintaining the expressiveness and naturalness of vocal performances. In BERT-APC, a novel stationary pitch predictor first estimates the stationary pitch of each note-defined as the continuous pitch derived from the stable region within a note and serving as an approximation for the perceived pitch-from the detuned singing voice. Subsequently, a context-aware note pitch predictor infers the intended pitch sequence by leveraging a repurposed music language model that incorporates musical context. Finally, a note-level correction algorithm fixes note-level pitch errors while preserving intentional pitch deviations for emotional expression. In addition, we introduce a learnable data augmentation strategy that improves the robustness of the music language model by simulating realistic detuning patterns. Compared to two recent singing voice transcription models, BERT-APC demonstrated superior performance in target note pitch prediction, outperforming the second-best model, ROSVOT, by 10.49 percentage points (pp) on highly detuned samples in terms of the raw pitch accuracy. In the MOS test, BERT-APC achieved the highest quality rating of 4.32 $\pm$ 0.15, which is significantly higher than those of the widely-used commercial APC tools, Auto-Tune (3.22 $\pm$ 0.18) and Melodyne (3.08 $\pm$ 0.18), while maintaining a comparable ability to preserve expressive nuances. To the best of our knowledge, this is the first APC model that leverages a music language model to achieve reference-free pitch correction with symbolic musical context. The corrected audio samples of BERT-APC are available online.
- 28 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 34, 2649-2662
BERT-APC: A Reference-Free Framework for Automatic Pitch Correction via Musical Context Inference
Automatic Pitch Correction (APC) enhances vocal recordings by correcting pitch deviations to align with the intended musical notes. However, existing APC systems either rely on reference pitches, which limits their practical applicability, or employ simple pitch estimation algorithms that often fail to preserve expressiveness and naturalness. We propose BERT-APC, a novel reference-free APC framework that corrects pitch errors while maintaining the expressiveness and naturalness of vocal performances. In BERT-APC, a novel stationary pitch predictor first estimates the stationary pitch of each note-defined as the continuous pitch derived from the stable region within a note and serving as an approximation for the perceived pitch-from the detuned singing voice. Subsequently, a context-aware note pitch predictor infers the intended pitch sequence by leveraging a repurposed music language model that incorporates musical context. Finally, a note-level correction algorithm fixes note-level pitch errors while preserving intentional pitch deviations for emotional expression. In addition, we introduce a learnable data augmentation strategy that improves the robustness of the music language model by simulating realistic detuning patterns. Compared to two recent singing voice transcription models, BERT-APC demonstrated superior performance in target note pitch prediction, outperforming the second-best model, ROSVOT, by 10.49 percentage points (pp) on highly detuned samples in terms of the raw pitch accuracy. In the MOS test, BERT-APC achieved the highest quality rating of 4.32 $\pm$ 0.15, which is significantly higher than those of the widely-used commercial APC tools, Auto-Tune (3.22 $\pm$ 0.18) and Melodyne (3.08 $\pm$ 0.18), while maintaining a comparable ability to preserve expressive nuances. To the best of our knowledge, this is the first APC model that leverages a music language model to achieve reference-free pitch correction with symbolic musical context. The corrected audio samples of BERT-APC are available online.
- 11 April 2026
- Published by Springer Nature in Neural Computing and Applications
- Vol. 38 (8)
Holi-DETR: holistic fashion item detection leveraging contextual information
- 11 April 2026
- Published by Springer Nature in Neural Computing and Applications
- Vol. 38 (8)
Holi-DETR: holistic fashion item detection leveraging contextual information
- Open Access
Rethinking Garment Conditioning in Diffusion-based Virtual Try-On
Virtual Try-On (VTON) is the task of synthesizing an image of a person wearing a target garment, conditioned on a person image and a garment image. While diffusion-based VTON models featuring a Dual UNet architecture demonstrate superior fidelity compared to single UNet models, they incur substantial computational and memory overhead due to their heavy structure. In this study, through visualization analysis and theoretical analysis, we derived three hypotheses regarding the learning of context features to condition the denoising process. Based on these hypotheses, we developed Re-CatVTON, an efficient single UNet model that achieves high performance. We further enhance the model by introducing a modified classifier-free guidance strategy tailored for VTON's spatial concatenation conditioning, and by directly injecting the ground-truth garment latent derived from the clean garment latent to prevent the accumulation of prediction error. The proposed Re-CatVTON significantly improves performance compared to its predecessor (CatVTON) and requires less computation and memory than the high-performance Dual UNet model, Leffa. Our results demonstrate improved FID, KID, and LPIPS scores, with only a marginal decrease in SSIM, establishing a new efficiency-performance trade-off for single UNet VTON models.
Open AccessRethinking Garment Conditioning in Diffusion-based Virtual Try-On
Virtual Try-On (VTON) is the task of synthesizing an image of a person wearing a target garment, conditioned on a person image and a garment image. While diffusion-based VTON models featuring a Dual UNet architecture demonstrate superior fidelity compared to single UNet models, they incur substantial computational and memory overhead due to their heavy structure. In this study, through visualization analysis and theoretical analysis, we derived three hypotheses regarding the learning of context features to condition the denoising process. Based on these hypotheses, we developed Re-CatVTON, an efficient single UNet model that achieves high performance. We further enhance the model by introducing a modified classifier-free guidance strategy tailored for VTON's spatial concatenation conditioning, and by directly injecting the ground-truth garment latent derived from the clean garment latent to prevent the accumulation of prediction error. The proposed Re-CatVTON significantly improves performance compared to its predecessor (CatVTON) and requires less computation and memory than the high-performance Dual UNet model, Leffa. Our results demonstrate improved FID, KID, and LPIPS scores, with only a marginal decrease in SSIM, establishing a new efficiency-performance trade-off for single UNet VTON models.
- Open Access
- 7 August 2025
- Published by IOS Press
- Vol. 329, 896-900
A New Approach to Perfusion-Weighted Image Analysis: Measurement via Red–Green–Blue Information Inversion
This study developed and validated a Python-based program for analyzing the region of interest mean values using red–green–blue (RGB)–reconstructed perfusion-weighted imaging (PWI) data. The program addressed the limitations of traditional source-image-based analysis, including manufacturer dependency, high storage costs, and limited accessibility. High intraclass correlation coefficient (ICC) values for cerebral blood volume and time to peak confirmed the feasibility of this approach, whereas lower ICC values for mean transit time suggested discrepancies between reconstructed RGB images and conventional dynamic susceptibility contrast–based data. Designed with a user-friendly graphical user interface and executable distribution, the program enhances clinical applicability. Future studies should focus on improving the algorithms and validating the approach with multicenter datasets.
- 7 August 2025
- Published by IOS Press
- Vol. 329, 896-900
Open AccessA New Approach to Perfusion-Weighted Image Analysis: Measurement via Red–Green–Blue Information Inversion
This study developed and validated a Python-based program for analyzing the region of interest mean values using red–green–blue (RGB)–reconstructed perfusion-weighted imaging (PWI) data. The program addressed the limitations of traditional source-image-based analysis, including manufacturer dependency, high storage costs, and limited accessibility. High intraclass correlation coefficient (ICC) values for cerebral blood volume and time to peak confirmed the feasibility of this approach, whereas lower ICC values for mean transit time suggested discrepancies between reconstructed RGB images and conventional dynamic susceptibility contrast–based data. Designed with a user-friendly graphical user interface and executable distribution, the program enhances clinical applicability. Future studies should focus on improving the algorithms and validating the approach with multicenter datasets.
- 5 April 2025
- Published by Elsevier in Computer Vision and Image Understanding
- Vol. 256
MF-LPR 2 : Multi-frame license plate image restoration and recognition using optical flow
- 5 April 2025
- Published by Elsevier in Computer Vision and Image Understanding
- Vol. 256
MF-LPR 2 : Multi-frame license plate image restoration and recognition using optical flow
- 18 July 2024
- Published by Springer Nature in Applied Intelligence
- Vol. 54 (20) , 9579-9593
Item-region-based style classification network (IRSN): a fashion style classifier based on domain knowledge of fashion experts
- 18 July 2024
- Published by Springer Nature in Applied Intelligence
- Vol. 54 (20) , 9579-9593
Item-region-based style classification network (IRSN): a fashion style classifier based on domain knowledge of fashion experts
- 29 February 2024
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 172, 108241
PECI-Net: Bolus segmentation from video fluoroscopic swallowing study images using preprocessing ensemble and cascaded inference
- 29 February 2024
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 172, 108241
PECI-Net: Bolus segmentation from video fluoroscopic swallowing study images using preprocessing ensemble and cascaded inference