analiza aiqbench
This research introduces Visual Oscillator Networks and an Attention Broadcast Decoder to create interpretable models for soft continuum robots directly from video data. By generating spatially grounded attention maps, the method reveals latent mechanical properties like mass distribution and stiffness without requiring manual prior knowledge or model design. The approach significantly improves multi-step prediction accuracy while enabling autonomous discovery of oscillator chain structures relevant for future control applications.
Learning Visually Interpretable Oscillator Networks for Soft Continuum Robots from Video
arXiv:2511.18322v4 Announce Type: replace Abstract: Learning soft continuum robot (SCR) dynamics from video offers flexibility but existing methods lack interpretability or rely on prior assumptions. Model-based approaches require prior knowledge and manual design. We bridge this gap by introducing: (1) The Attention Broadcast Decoder (ABCD), a plug-and-play module for autoencoder-based latent dynamics learning that generates pixel-accurate attention maps localizing each latent dimension's contribution while filtering static backgrounds, enabling visual interpretability via spatially grounded latents and on-image overlays. (2) Visual Oscillator Networks (VONs), a 2D latent oscillator network coupled to ABCD attention maps for on-image visualization of learned masses, coupling stiffness, and forces, thereby enabling mechanical interpretability. We validate our approach on single- and double-segment SCRs, demonstrating that ABCD-based models significantly improve multi-step prediction accuracy with 5.8x error reduction for Koopman operators and 3.5x for oscillator networks on a two-segment robot. VONs autonomously discover a chain structure of oscillators. This fully data-driven approach yields compact, mechanically interpretable models with potential relevance for future control applications.
Abstract:Learning soft continuum robot (SCR) dynamics from video offers flexibility but existing methods lack interpretability or rely on prior assumptions. Model-based approaches require prior knowledge and manual design. We bridge this gap by introducing: (1) The Attention Broadcast Decoder (ABCD), a plug-and-play module for autoencoder-based latent dynamics learning that generates pixel-accurate attention maps localizing each latent dimension's contribution while filtering static backgrounds, enabling visual interpretability via spatially grounded latents and on-image overlays. (2) Visual Oscillator Networks (VONs), a 2D latent oscillator network coupled to ABCD attention maps for on-image visualization of learned masses, coupling stiffness, and forces, thereby enabling mechanical interpretability. We validate our approach on single- and double-segment SCRs, demonstrating that ABCD-based models significantly improve multi-step prediction accuracy with 5.8x error reduction for Koopman operators and 3.5x for oscillator networks on a two-segment robot. VONs autonomously discover a chain structure of oscillators. This fully data-driven approach yields compact, mechanically interpretable models with potential relevance for future control applications.
Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
arXiv:2511.18322 [cs.RO]
(or arXiv:2511.18322v4 [cs.RO] for this version)
Text integral sub licența deschisă a publicației · 178 cuvinte
Subiecte conexe
Accurate Open-Loop Control of a Soft Continuum Robot Using Visually Learned Latent Dynamics
arXiv · 5 oct. 2026
Humanity’s Last Invention — Richard Socher of Recursive
Latent Space · 14 sept. 2026
Evaluating task vectors, unlearning and inoculation
LessWrong · 20 sept. 2026
Coming of a New Sun
LessWrong · 3 aug. 2026