arXiv subjects

Seulgi Jeong

Publications and source records attributed to Seulgi Jeong.

3 recordsLinked to original sources

arXiv2026

H-Adapter: Pose-Robust Hairstyle Transfer via Attention-Derived, Source-Aligned Hair Masks

Hairstyle transfer has practical applications such as virtual try-on, yet remains challenging when the source and reference exhibit large head-pose discrepancies. We propose H-Adapter, which improves pose robustness by training with a region-specific loss that disentangles hair and non-hair objectives and thereby induces spatially disentangled cross-attention, from which a source-aligned hair edit mask is derived to guide diffusion-based inpainting. Experiments on pose-agnostic and pose-different subsets demonstrate strong quantitative results, including the best FID, $\mathrm{FID}_{\mathrm{CLIP}}$, and CLIP-I under pose differences, while maintaining competitive non-hair preservation and improving qualitative fidelity to fine-grained reference hairstyle details. Beyond source-conditioned transfer, H-Adapter supports practical extensions including text-to-image generation, auxiliary prompt-based hair color control, and compatibility with an identity-preserving IP-Adapter variant. We also introduce a VLM-as-a-judge protocol and observe consistent gains in hairstyle faithfulness, non-hair preservation, and artifact quality.

cs.CV↗

arXiv2025

MINDiff: Mask-Integrated Negative Attention for Controlling Overfitting in Text-to-Image Personalization

In the personalization process of large-scale text-to-image models, overfitting often occurs when learning specific subject from a limited number of images. Existing methods, such as DreamBooth, mitigate this issue through a class-specific prior-preservation loss, which requires increased computational cost during training and limits user control during inference time. To address these limitations, we propose Mask-Integrated Negative Attention Diffusion (MINDiff). MINDiff introduces a novel concept, negative attention, which suppresses the subject's influence in masked irrelevant regions. We achieve this by modifying the cross-attention mechanism during inference. This enables semantic control and improves text alignment by reducing subject dominance in irrelevant regions. Additionally, during the inference time, users can adjust a scale parameter lambda to balance subject fidelity and text alignment. Our qualitative and quantitative experiments on DreamBooth models demonstrate that MINDiff mitigates overfitting more effectively than class-specific prior-preservation loss. As our method operates entirely at inference time and does not alter the model architecture, it can be directly applied to existing DreamBooth models without re-training. Our code is available at https://github.com/seuleepy/MINDiff.

cs.CV↗

arXiv2013

On the semi-threading of knot diagrams with minimal overpasses

Given a knot diagram $D$, we construct a semi-threading circle for it which can be an axis of $D$ as a closed braid depending on knot diagrams. In particular, we consider semi-threading circles for minimal diagrams of a knot with respect to overpasses which give us some information related to the braid index. By this notion, we show that, for every nontrivial knot $K$, the braid index $b(K)$ of $K$ is not less than the minimum number $l(K)$ of overpasses of diagrams. Moreover, they are the same for a torus knot.

math.GN↗