LanPerAct: A Framework for Language-Driven Perception and Robotic Manipulation
Bibliographic record
Abstract
Robotic manipulation necessitates comprehensive environmental perception and precise awareness of object pose estimation. Current systems predominantly rely on predefined motion primitives, exhibiting significant limitations in flexible interaction. The semantic reasoning capabilities of Large Language Models can provide robust task planning and decision-making support, while more accurate object pose estimation ensures execution precision and operational stability. In this study, we present LanPerAct, a Unified Framework for Language-Driven Perception and Robotic Manipulation. LanPerAct integrates large language models with pose-based grasping through a language-vision-action co-reasoning pipeline. Our approach first employs Deepseek to construct an end-to-end natural language instruction parser that generates structured task parameters. Subsequently, we utilize open-vocabulary object detection and instance segmentation to obtain object masks and corresponding point cloud data, followed by a two-stage point-matching algorithm for precise 6D pose estimation. Finally, the system executes zero-shot object grasping through generated code that interfaces with our multi-stage, multi-task pose-grasping framework, achieving both lightweight implementation and robust task execution. Experimental results in both simulated and real-world environments demonstrate the advanced effectiveness of our method, showcasing its capability for efficient and precise execution of natural language-guided manipulation tasks. We believe this research provides valuable insights for real-world robotic behavior development.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.005 | 0.005 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.008 | 0.004 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".