Sunghoon Im
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Sunghoon Im has a total of 87 co-authors. Here are the top co-authors by number of shared publications:
- Jaeyeul Kim
Daegu Gyeongbuk Institute of Science and Technology (DGIST)
8 shared publications - Hae-Gon Jeon
Department of Artificial Intelligence, Yonsei University, Seoul, South Korea
7 shared publications - In So Kweon
KAIST
5 shared publications - Jihun Park
Daegu Gyeongbuk Institute of Science and Technology, Daegu, Republic of Korea
4 shared publications - Jungwan Woo
Department of Electrical Engineering and Computer Science, DGIST, Daegu, Republic of Korea
4 shared publications - Kyumin Hwang
Department of Electrical Engineering & Computer Sciences Daegu Gyeongbuk Institute of Science and Technology (DGIST) 42988 Daegu South Korea
4 shared publications
Latest Publications
- Open Access
- 30 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-11
CRAFT: Constrained Reward via Attention Fine-Tuning for Subject Personalization without Composed Targets
- 30 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Open AccessCRAFT: Constrained Reward via Attention Fine-Tuning for Subject Personalization without Composed Targets
- 10 September 2026
- Published by Springer Nature
- p. 673-692
Reference-to-Image (R2I) Aware Latent Diffusion Model for Image Super-Resolution
- 10 September 2026
- Published by Springer Nature
- p. 673-692
Reference-to-Image (R2I) Aware Latent Diffusion Model for Image Super-Resolution
- Open Access
- 7 September 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Multimedia Computing, Communications, and Applications
- Vol. 22 (9) , 1-20
Mitigating Noisy Correspondence in Video–Text Retrieval via Noise-Mined Adaptive Self-Labeling
In video–text retrieval, addressing the noisy correspondence problem is crucial for achieving accurate retrieval performance. Recent methods address this challenge by either suppressing the impact of noisy data or predictions distorted by noise. However, existing approaches often overlook two critical aspects: distinguishing hard noise from semantically ambiguous cases and preserving latent associations within unmatched negative pairs. To address this limitation, we propose Noise-mined Adaptive Self-Labeling (NASL), which effectively manages noisy data during training. NASL consists of two loss functions: (1) Noise-mined Matching Loss (NML), which identifies and penalizes noisy data based on a two-stage suppression strategy, and (2) Adaptive Self-labeling Loss (ASL), which employs optimal transport to recover latent associations among false negatives in noisy conditions and provides soft supervision to prevent excessive penalization of semantically plausible pairs. Extensive experiments demonstrate that NASL improves the separation between clean and noisy data while effectively mitigating noise, leading to significant performance improvements on the MSR-VTT, DiDeMo, MSVD, LSMDC, and ActivityNet datasets.
- 7 September 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Multimedia Computing, Communications, and Applications
- Vol. 22 (9) , 1-20
Open AccessMitigating Noisy Correspondence in Video–Text Retrieval via Noise-Mined Adaptive Self-Labeling
In video–text retrieval, addressing the noisy correspondence problem is crucial for achieving accurate retrieval performance. Recent methods address this challenge by either suppressing the impact of noisy data or predictions distorted by noise. However, existing approaches often overlook two critical aspects: distinguishing hard noise from semantically ambiguous cases and preserving latent associations within unmatched negative pairs. To address this limitation, we propose Noise-mined Adaptive Self-Labeling (NASL), which effectively manages noisy data during training. NASL consists of two loss functions: (1) Noise-mined Matching Loss (NML), which identifies and penalizes noisy data based on a two-stage suppression strategy, and (2) Adaptive Self-labeling Loss (ASL), which employs optimal transport to recover latent associations among false negatives in noisy conditions and provides soft supervision to prevent excessive penalization of semantically plausible pairs. Extensive experiments demonstrate that NASL improves the separation between clean and noisy data while effectively mitigating noise, leading to significant performance improvements on the MSR-VTT, DiDeMo, MSVD, LSMDC, and ActivityNet datasets.
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 13259-13264
Learning to Forget: Emotional Salience as a Compression Mechanism for Long-Term AI Memory
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 13259-13264
Open AccessLearning to Forget: Emotional Salience as a Compression Mechanism for Long-Term AI Memory
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33, 3586-3590
Object-Centric LiDAR-to-Radar Distillation for 3D Object Detection
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33, 3586-3590
Object-Centric LiDAR-to-Radar Distillation for 3D Object Detection
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33 (10709908) , 1506-1510
CascadeOcc: Rethinking 3D Occupancy World Models With Cascaded VQ Representations
This letter proposes CascadeOcc, a novel occupancy world model that prioritizes intrinsic structural hierarchy over extrinsic auxiliary modalities for autonomous driving. Occupancy world models—forecasting the future driving environment and planning the driving trajectory—effectively bridge perception and planning, but current approaches often heavily rely on external modalities or large language models, failing to fully exploit the inherent structural potential of occupancy representations themselves. To enhance representational capacity for complex 3D scenes, we integrate a cascaded Vector Quantized (VQ) mechanism into an autoregressive framework. Following a coarse-to-fine principle, CascadeOcc progressively refines fine-grained details from global structures through a multi-scale architecture. Additionally, we incorporate a TimeMixer to capture multi-scale temporal dependencies, establishing a dual-hierarchy mechanism in both space and time. Experimental results on 4D occupancy forecasting and motion planning benchmarks demonstrate that CascadeOcc achieves superior performance among vision-centric approaches, validating that optimizing inherent representations is a powerful alternative to relying on external foundation models.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33 (10709908) , 1506-1510
CascadeOcc: Rethinking 3D Occupancy World Models With Cascaded VQ Representations
This letter proposes CascadeOcc, a novel occupancy world model that prioritizes intrinsic structural hierarchy over extrinsic auxiliary modalities for autonomous driving. Occupancy world models—forecasting the future driving environment and planning the driving trajectory—effectively bridge perception and planning, but current approaches often heavily rely on external modalities or large language models, failing to fully exploit the inherent structural potential of occupancy representations themselves. To enhance representational capacity for complex 3D scenes, we integrate a cascaded Vector Quantized (VQ) mechanism into an autoregressive framework. Following a coarse-to-fine principle, CascadeOcc progressively refines fine-grained details from global structures through a multi-scale architecture. Additionally, we incorporate a TimeMixer to capture multi-scale temporal dependencies, establishing a dual-hierarchy mechanism in both space and time. Experimental results on 4D occupancy forecasting and motion planning benchmarks demonstrate that CascadeOcc achieves superior performance among vision-centric approaches, validating that optimizing inherent representations is a powerful alternative to relying on external foundation models.
- Open Access
Scale-invariant and View-relational Representation Learning for Full Surround Monocular Depth
Recent foundation models demonstrate strong generalization capabilities in monocular depth estimation. However, directly applying these models to Full Surround Monocular Depth Estimation (FSMDE) presents two major challenges: (1) high computational cost, which limits real-time performance, and (2) difficulty in estimating metric-scale depth, as these models are typically trained to predict only relative depth. To address these limitations, we propose a novel knowledge distillation strategy that transfers robust depth knowledge from a foundation model to a lightweight FSMDE network. Our approach leverages a hybrid regression framework combining the knowledge distillation scheme--traditionally used in classification--with a depth binning module to enhance scale consistency. Specifically, we introduce a cross-interaction knowledge distillation scheme that distills the scale-invariant depth bin probabilities of a foundation model into the student network while guiding it to infer metric-scale depth bin centers from ground-truth depth. Furthermore, we propose view-relational knowledge distillation, which encodes structural relationships among adjacent camera views and transfers them to enhance cross-view depth consistency. Experiments on DDAD and nuScenes demonstrate the effectiveness of our method compared to conventional supervised methods and existing knowledge distillation approaches. Moreover, our method achieves a favorable trade-off between performance and efficiency, meeting real-time requirements.
Open AccessScale-invariant and View-relational Representation Learning for Full Surround Monocular Depth
Recent foundation models demonstrate strong generalization capabilities in monocular depth estimation. However, directly applying these models to Full Surround Monocular Depth Estimation (FSMDE) presents two major challenges: (1) high computational cost, which limits real-time performance, and (2) difficulty in estimating metric-scale depth, as these models are typically trained to predict only relative depth. To address these limitations, we propose a novel knowledge distillation strategy that transfers robust depth knowledge from a foundation model to a lightweight FSMDE network. Our approach leverages a hybrid regression framework combining the knowledge distillation scheme--traditionally used in classification--with a depth binning module to enhance scale consistency. Specifically, we introduce a cross-interaction knowledge distillation scheme that distills the scale-invariant depth bin probabilities of a foundation model into the student network while guiding it to infer metric-scale depth bin centers from ground-truth depth. Furthermore, we propose view-relational knowledge distillation, which encodes structural relationships among adjacent camera views and transfers them to enhance cross-view depth consistency. Experiments on DDAD and nuScenes demonstrate the effectiveness of our method compared to conventional supervised methods and existing knowledge distillation approaches. Moreover, our method achieves a favorable trade-off between performance and efficiency, meeting real-time requirements.
- 21 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (1) , 1002-1009
Scale-Invariant and View-Relational Representation Learning for Full Surround Monocular Depth
Recent foundation models demonstrate strong generalization capabilities in monocular depth estimation. However, directly applying these models to Full Surround Monocular Depth Estimation (FSMDE) presentstwo major challenges: (1) high computational cost, which limits real-time performance, and (2) difficulty in estimating metric-scale depth, as these models are typically trained to predict only relative depth. To address these limitations, we propose a novel knowledge distillation strategy that transfers robust depth knowledge from a foundation model to a lightweight FSMDE network. Our approach leverages a hybrid regression framework combining the knowledge distillation scheme–traditionally used in classification–with a depth binning module to enhance scale consistency. Specifically, we introduce a cross-interaction knowledge distillation scheme that distills the scale-invariant depth bin probabilities of a foundation model into the student network while guiding it to infer metric-scale depth bin centers from ground-truth depth. Furthermore, we propose view-relational knowledge distillation, which encodes structural relationships among adjacent camera views and transfers them to enhance cross-view depth consistency. Experiments on DDAD and nuScenes demonstrate the effectiveness of our method compared to conventional supervised methods and existing knowledge distillation approaches. Moreover, our method achieves a favorable trade-off between performance and efficiency, meeting real-time requirements.
- 21 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (1) , 1002-1009
Scale-Invariant and View-Relational Representation Learning for Full Surround Monocular Depth
Recent foundation models demonstrate strong generalization capabilities in monocular depth estimation. However, directly applying these models to Full Surround Monocular Depth Estimation (FSMDE) presentstwo major challenges: (1) high computational cost, which limits real-time performance, and (2) difficulty in estimating metric-scale depth, as these models are typically trained to predict only relative depth. To address these limitations, we propose a novel knowledge distillation strategy that transfers robust depth knowledge from a foundation model to a lightweight FSMDE network. Our approach leverages a hybrid regression framework combining the knowledge distillation scheme–traditionally used in classification–with a depth binning module to enhance scale consistency. Specifically, we introduce a cross-interaction knowledge distillation scheme that distills the scale-invariant depth bin probabilities of a foundation model into the student network while guiding it to infer metric-scale depth bin centers from ground-truth depth. Furthermore, we propose view-relational knowledge distillation, which encodes structural relationships among adjacent camera views and transfers them to enhance cross-view depth consistency. Experiments on DDAD and nuScenes demonstrate the effectiveness of our method compared to conventional supervised methods and existing knowledge distillation approaches. Moreover, our method achieves a favorable trade-off between performance and efficiency, meeting real-time requirements.
- Open Access
Infinite-Story: A Training-Free Consistent Text-to-Image Generation
We present Infinite-Story, a training-free framework for consistent text-to-image (T2I) generation tailored for multi-prompt storytelling scenarios. Built upon a scale-wise autoregressive model, our method addresses two key challenges in consistent T2I generation: identity inconsistency and style inconsistency. To overcome these issues, we introduce three complementary techniques: Identity Prompt Replacement, which mitigates context bias in text encoders to align identity attributes across prompts; and a unified attention guidance mechanism comprising Adaptive Style Injection and Synchronized Guidance Adaptation, which jointly enforce global style and identity appearance consistency while preserving prompt fidelity. Unlike prior diffusion-based approaches that require fine-tuning or suffer from slow inference, Infinite-Story operates entirely at test time, delivering high identity and style consistency across diverse prompts. Extensive experiments demonstrate that our method achieves state-of-the-art generation performance, while offering over 6X faster inference (1.72 seconds per image) than the existing fastest consistent T2I models, highlighting its effectiveness and practicality for real-world visual storytelling.
Open AccessInfinite-Story: A Training-Free Consistent Text-to-Image Generation
We present Infinite-Story, a training-free framework for consistent text-to-image (T2I) generation tailored for multi-prompt storytelling scenarios. Built upon a scale-wise autoregressive model, our method addresses two key challenges in consistent T2I generation: identity inconsistency and style inconsistency. To overcome these issues, we introduce three complementary techniques: Identity Prompt Replacement, which mitigates context bias in text encoders to align identity attributes across prompts; and a unified attention guidance mechanism comprising Adaptive Style Injection and Synchronized Guidance Adaptation, which jointly enforce global style and identity appearance consistency while preserving prompt fidelity. Unlike prior diffusion-based approaches that require fine-tuning or suffer from slow inference, Infinite-Story operates entirely at test time, delivering high identity and style consistency across diverse prompts. Extensive experiments demonstrate that our method achieves state-of-the-art generation performance, while offering over 6X faster inference (1.72 seconds per image) than the existing fastest consistent T2I models, highlighting its effectiveness and practicality for real-world visual storytelling.
- Open Access
JPEG Processing Neural Operator for Backward-Compatible Coding
Despite significant advances in learning-based lossy compression algorithms, standardizing codecs remains a critical challenge. In this paper, we present the JPEG Processing Neural Operator (JPNeO), a next-generation JPEG algorithm that maintains full backward compatibility with the current JPEG format. Our JPNeO improves chroma component preservation and enhances reconstruction fidelity compared to existing artifact removal methods by incorporating neural operators in both the encoding and decoding stages. JPNeO achieves practical benefits in terms of reduced memory usage and parameter count. We further validate our hypothesis about the existence of a space with high mutual information through empirical evidence. In summary, the JPNeO functions as a high-performance out-of-the-box image compression pipeline without changing source coding's protocol. Our source code is available at https://github.com/WooKyoungHan/JPNeO.
Open AccessJPEG Processing Neural Operator for Backward-Compatible Coding
Despite significant advances in learning-based lossy compression algorithms, standardizing codecs remains a critical challenge. In this paper, we present the JPEG Processing Neural Operator (JPNeO), a next-generation JPEG algorithm that maintains full backward compatibility with the current JPEG format. Our JPNeO improves chroma component preservation and enhances reconstruction fidelity compared to existing artifact removal methods by incorporating neural operators in both the encoding and decoding stages. JPNeO achieves practical benefits in terms of reduced memory usage and parameter count. We further validate our hypothesis about the existence of a space with high mutual information through empirical evidence. In summary, the JPNeO functions as a high-performance out-of-the-box image compression pipeline without changing source coding's protocol. Our source code is available at https://github.com/WooKyoungHan/JPNeO.