@misc{indiciaeaad85bd44341,
  title = {Unlocking Fine-Grained Perception in CLIP via Structurally-Aware Latent Masked Modeling},
  author = {Juntong Li and Lingwei Dang and Haomin Wu and Ziyan Qiu and Qingxin Xiao and Qingyao Wu},
  year = {2026},
  url = {https://arxiv.org/abs/2610.07689},
  note = {Source identifier: 2610.07689}
}