@misc{indiciae3e5db0ef761e, title = {VisualErase: Dual-Branch Visual Trajectory Redirection for Robust Concept Erasure in Text-to-Image Diffusion Models}, author = {Qianlong Xiang and Miao Zhang and Kun Wang and Yupeng Hu and Junhui Hou and Liqiang Nie}, year = {2026}, url = {https://arxiv.org/abs/2610.05000}, note = {Source identifier: 2610.05000} }