From Observation to Action: Latent Action-based Primitive Segmentation for VLA Pre-training in Industrial Settings

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Jiajie, Schwertfeger, Sören, Kleiner, Alexander
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915897846792192
author Zhang, Jiajie
Schwertfeger, Sören
Kleiner, Alexander
author_facet Zhang, Jiajie
Schwertfeger, Sören
Kleiner, Alexander
contents We present a novel unsupervised framework to unlock vast unlabeled human demonstration data from continuous industrial video streams for Vision-Language-Action (VLA) model pre-training. Our method first trains a lightweight motion tokenizer to encode motion dynamics, then employs an unsupervised action segmenter leveraging a novel "Latent Action Energy" metric to discover and segment semantically coherent action primitives. The pipeline outputs both segmented video clips and their corresponding latent action sequences, providing structured data directly suitable for VLA pre-training. Evaluations on public benchmarks and a proprietary electric motor assembly dataset demonstrate effective segmentation of key tasks performed by humans at workstations. Further clustering and quantitative assessment via a Vision-Language Model confirm the semantic coherence of the discovered action primitives. To our knowledge, this is the first fully automated end-to-end system for extracting and organizing VLA pre-training data from unstructured industrial videos, offering a scalable solution for embodied AI integration in manufacturing.
format Preprint
id arxiv_https___arxiv_org_abs_2511_21428
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle From Observation to Action: Latent Action-based Primitive Segmentation for VLA Pre-training in Industrial Settings
Zhang, Jiajie
Schwertfeger, Sören
Kleiner, Alexander
Computer Vision and Pattern Recognition
Artificial Intelligence
Robotics
We present a novel unsupervised framework to unlock vast unlabeled human demonstration data from continuous industrial video streams for Vision-Language-Action (VLA) model pre-training. Our method first trains a lightweight motion tokenizer to encode motion dynamics, then employs an unsupervised action segmenter leveraging a novel "Latent Action Energy" metric to discover and segment semantically coherent action primitives. The pipeline outputs both segmented video clips and their corresponding latent action sequences, providing structured data directly suitable for VLA pre-training. Evaluations on public benchmarks and a proprietary electric motor assembly dataset demonstrate effective segmentation of key tasks performed by humans at workstations. Further clustering and quantitative assessment via a Vision-Language Model confirm the semantic coherence of the discovered action primitives. To our knowledge, this is the first fully automated end-to-end system for extracting and organizing VLA pre-training data from unstructured industrial videos, offering a scalable solution for embodied AI integration in manufacturing.
title From Observation to Action: Latent Action-based Primitive Segmentation for VLA Pre-training in Industrial Settings
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Robotics
url https://arxiv.org/abs/2511.21428