arXiv:2501. 12289v3 Announce Type: replace-cross Abstract: Internet overuse is a widespread phenomenon in today's digital society.
Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.
Emotion-driven image editing aims to evoke a specified target emotion by modifying emotion-relevant visual cues in a source image, while preserving the overall composition and semantic-structural cohe...
arXiv:2609.02616v1 Announce Type: new Abstract: Emotion-driven image editing aims to evoke a specified target emotion by modifying emotion-relevant visual cues in a source image, while preserving the...
arXiv:2606. 09901v1 Announce Type: cross Abstract: Diffusion-based generative models enable powerful image editing capabilities, but achieving precise control while maintaining fidelity and safety remains challenging.
arXiv:2605. 31162v1 Announce Type: cross Abstract: Unconditional diffusion models offer powerful generative priors, yet steering them toward aesthetically enhanced outputs remains largely unexplored.
arXiv:2608.28802v1 Announce Type: cross Abstract: Facial appearance editing powers popular applications like FaceApp and Photoshop. Generative Adversarial Networks (GANs) and 3D Morphable Models (3DM...
arXiv:2606. 13247v1 Announce Type: new Abstract: Text-to-image diffusion models have achieved impressive results in synthesizing high-quality images from natural language prompts.