ProcVLM: Learning Procedure-Grounded Progress Rewards for Robotic Manipulation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915997458366464 |
|---|---|
| author | Feng, Youhe Shi, Hansen Li, Haoyang Guo, Xinlei Wang, Yang Zhang, Chengyang Zhang, Jinkai Zhang, Xiaohan Tang, Jie Zhang, Jing |
| author_facet | Feng, Youhe Shi, Hansen Li, Haoyang Guo, Xinlei Wang, Yang Zhang, Chengyang Zhang, Jinkai Zhang, Xiaohan Tang, Jie Zhang, Jing |
| contents | Long-horizon robotic manipulation requires dense feedback that reflects how a task advances through its procedural stages, not merely whether the final outcome is successful. Existing reward models often rely on trajectory-level success labels or time-based interpolation, which can conflate elapsed time with true task progress and therefore fail to capture unfinished steps, stagnation, and failure states. We present ProcVLM, a progress-aware vision-language model that learns procedure-grounded progress as a dense reward signal for manipulation. Rather than deriving progress from terminal outcomes or temporal proxies, ProcVLM grounds progress estimation in procedural structure and intra-stage visual change, and further adopts a reasoning-before-estimation paradigm that infers the remaining atomic actions before estimating task progress. Specifically, we construct this supervision by synthesizing frame-level subtask-semantic annotations, assigning progress budgets according to subtask structure, and distributing each budget based on intra-subtask visual change. To train ProcVLM at scale, we build a standardized procedural supervision synthesis pipeline and construct ProcCorpus-60M from 30 embodied datasets with 60M annotated frames, from which we derive ProcVQA for procedure-aware pretraining, with progress estimation as the central task alongside action segmentation and future planning. Experiments on ProcVQA and reward-model benchmarks show that ProcVLM improves embodied procedural reasoning and yields more discriminative trajectory-internal progress estimates than representative baselines, supporting its use as a dense reward model for downstream reward-guided policy optimization. Project page: https://procvlm.github.io/ |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_08774 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | ProcVLM: Learning Procedure-Grounded Progress Rewards for Robotic Manipulation Feng, Youhe Shi, Hansen Li, Haoyang Guo, Xinlei Wang, Yang Zhang, Chengyang Zhang, Jinkai Zhang, Xiaohan Tang, Jie Zhang, Jing Robotics Machine Learning Long-horizon robotic manipulation requires dense feedback that reflects how a task advances through its procedural stages, not merely whether the final outcome is successful. Existing reward models often rely on trajectory-level success labels or time-based interpolation, which can conflate elapsed time with true task progress and therefore fail to capture unfinished steps, stagnation, and failure states. We present ProcVLM, a progress-aware vision-language model that learns procedure-grounded progress as a dense reward signal for manipulation. Rather than deriving progress from terminal outcomes or temporal proxies, ProcVLM grounds progress estimation in procedural structure and intra-stage visual change, and further adopts a reasoning-before-estimation paradigm that infers the remaining atomic actions before estimating task progress. Specifically, we construct this supervision by synthesizing frame-level subtask-semantic annotations, assigning progress budgets according to subtask structure, and distributing each budget based on intra-subtask visual change. To train ProcVLM at scale, we build a standardized procedural supervision synthesis pipeline and construct ProcCorpus-60M from 30 embodied datasets with 60M annotated frames, from which we derive ProcVQA for procedure-aware pretraining, with progress estimation as the central task alongside action segmentation and future planning. Experiments on ProcVQA and reward-model benchmarks show that ProcVLM improves embodied procedural reasoning and yields more discriminative trajectory-internal progress estimates than representative baselines, supporting its use as a dense reward model for downstream reward-guided policy optimization. Project page: https://procvlm.github.io/ |
| title | ProcVLM: Learning Procedure-Grounded Progress Rewards for Robotic Manipulation |
| topic | Robotics Machine Learning |
| url | https://arxiv.org/abs/2605.08774 |