Skip to content
Preprint

AnyviewMeter: Adapting Robotic Reward Models with Camera Geometry and Multi-View Attention

Sep 2026 · 0 citations · 35 references
Computer Science

TL;DR

AnyviewMeter is introduced, a geometry-conditioned adaptation framework for robotic reward models that represent task progress as a scalar reward signal that combines low-rank fine-tuning with token-aligned Plucker rays and synchronous block attention.

Abstract

Robotic reward models evaluate task execution from visual observations, but their predictions can change with camera viewpoint and occlusion even when the underlying task state is unchanged. Adapting a pretrained reward model to a local task therefore requires accounting for how that task is observed. We introduce AnyviewMeter, a geometry-conditioned adaptation framework for robotic reward models that represent task progress as a scalar reward signal. It combines low-rank fine-tuning with token-aligned Plucker rays and synchronous block attention: ray conditioning incorporates camera geometry into visual features and attention queries and keys, while block attention fuses synchronized views inside the pretrained decoder. The framework supports both single-view reward prediction and joint multi-view evaluation through parameter-efficient adaptation of a pretrained Robometer model. On PickCube, single-view adaptation improves progress prediction in every camera group and reduces mean absolute error under a changed field of view by approximately 21% relative to RGB fine-tuning. Across simulated manipulation tasks, joint multi-view prediction reduces progress error by 41-69% compared with averaging single-view RGB predictions and improves temporal ordering in approximately 88% of task-camera groups. On real tasks with fixed and wrist-mounted cameras, mean absolute error decreases by approximately 21% relative to averaged RGB fine-tuning. These results support camera geometry and joint visual evidence as useful components of task-specific robotic reward adaptation.

View source

Similar papers

#artificial intelligence Preprint Oct 2026

CriticHack: Evaluating Visual Rewards Under Robot Policy Optimization

Learned visual reward models are increasingly used to optimize robot policies, yet a reward model can score an execution that acts on the wrong object as highly as one that completes the task. We show that optimizing such a reward can amplify these wrong-object failures while reward and task success both rise, so the s...

Jia-Xuan Luo, Xing-Guo Xu, Shan-Shan Wang et al. · 2 citations
Review Sep 2026

WorldReward: Reward Modeling for Camera-Conditioned World Models

This work presents WorldReward, a VLM-based pairwise preference reward model that unifies action-consistency and visual-quality evaluation for camera-conditioned world models, and introduces WorldReward-Bench, a human-annotated benchmark measuring reward-model agreement with human preferences across action consistency,...

Yi-Bin Wang, Ze-Han Wang, Junshu Tang et al. · 0 citations
Preprint Oct 2026

ActiveWAM: Evidence-Aware Active Vision for World-Action Models

Active vision manipulation requires a policy to control both its camera and its end-effectors, yet camera motion determines which evidence remains visible within finite observation windows. Acquiring a new view can displace task-critical cues, while retaining a view forgoes potentially useful observations. We formulate...

Ren-Jun Wu, Lu-Zhou Ge, Xue-Song Li · 0 citations

ActiveScale: Scaling Active Perception for Robots across Model, Data, and Hardware

This work introduces ActiveScale, a framework that advances active perception through coordinated model, data, and hardware designs, and introduces Active-perception Mobile-manipulation Platform (AMP), a robotic platform that supports active perception and mobile manipulation through single-operator teleoperation, enab...

Shuai Zhou, Kai-Sheng Pang, Wen-Xuan Song et al. · 1 citation
Preprint Sep 2026

Recovering the View: Benchmarking Physical Active Vision for Occlusion Recovery in Robotic Manipulation

Physical active vision allows robots to change their viewpoint when task-relevant observations become unreliable, yet existing manipulation benchmarks provide limited support for studying how policies recover from occlusion during execution. We introduce BAVO-Bench (Bimanual Active Vision under Occlusion), a bimanual a...

Kai-Jun Luo, Yu-Di Huang, Qi-Jun Zhong et al. · 0 citations
#artificial intelligence Preprint Oct 2026

LoGo: Local-Global Rewards for Consistent Long-Horizon Video Generation

LoGo is introduced, which blends global and spatially localized rewards for camera-controlled video models, and effectively reduces local object shifts, artifacts, and global scene changes, illustrating the importance of credit assignment in post-training video models.

Ziqi Ma, Shreya Sharma, Mohamed El Banani et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.