Metraipolis › archive › dc_02f9af936ecb
TLC-DiT: Task-Aligned Local Visual Conditioning for Robust Multitask Robot Manipulation — arXiv:2609.34297
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.34297
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T00:17:21.609Z
- SHA-256 of the archived text
9f2b0e1fdf29feac89c4f1be2a1021fd8731df85c7530c92cd6ee9e6a4750d27- Stable URL of this page
https://metraipolis.com/documento/dc_02f9af936ecb
How to cite
TLC-DiT: Task-Aligned Local Visual Conditioning for Robust Multitask Robot Manipulation — arXiv:2609.34297 — arXiv (en), https://arxiv.org/abs/2609.34297 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_02f9af936ecb (Metraipolis archive dc_02f9af936ecb, sha256 9f2b0e1fdf29feac89c4f1be2a1021fd8731df85c7530c92cd6ee9e6a4750d27)
BibTeX
@misc{dc02f9af936ecb,
title = {TLC-DiT: Task-Aligned Local Visual Conditioning for Robust Multitask Robot Manipulation — arXiv:2609.34297},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.34297},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_02f9af936ecb; sha256 9f2b0e1fdf29feac89c4f1be2a1021fd8731df85c7530c92cd6ee9e6a4750d27},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.34297. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Language-conditioned robot policies have made clear progress in multitask manipulation, but task-relevant local visual evidence usually stays hidden inside a visual backbone or attention layers. This leaves the policy difficult to inspect and fragile under visual change, two symptoms of a missing explicit, task-aligned local visual channel. We present TLC-DiT, a plug-in extension of the Multitask Diffusion Transformer (DiT) policy that adds explicit task-guided local visual feature maps without changing the diffusion objective or the action-generation process. For each camera view, frozen DINOv2 patch features are modulated by the CLIP task embedding through FiLM and refined by a lightweight CoordConv CNN adapter into smooth spatial maps, which are concatenated with the original global image, language, joint-state, and timestep conditions. On LIBERO, TLC-DiT reaches a 93.5% average success rate, compared with 86.5% for Multitask DiT and 79.25% for SmolVLA. On LIBERO-plus, the total success rate improves from 54.07% to 57.24%, with larger gains under camera, background, and sensor-noise changes. In real-world bimanual tasks, TLC-DiT raises Teabag Putting completion from 44% to 89% while maintaining comparable Match Box Opening performance. Feature-map visualizations confirm that the model attends to task-relevant regions across views and perturbations, providing a direct way to inspect the visual evidence. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-29 in cs.RO). Categorie arXiv: Robotics (cs.RO). Pagina del lavoro: https://arxiv.org/abs/2609.34297 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-07T23:52:16.538Z dalla pagina https://arxiv.org/catchup/cs.RO/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).