Vision agents improve scientific image analysis by separating reasoning from action
Route First, See Later: Enabling Vision Agents to Reason Scientifically by Separating Routing Policy from Action Spaces
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
When computers analyze scientific images, they have to choose different tasks like detecting objects or counting them. The problem is that current systems mix up deciding what to do with actually doing it, making it hard to tell if mistakes come from poor decisions or from errors in image processing. The authors created a new system called ZETA that separates the decision of what task to do from performing the task itself. This lets users see exactly where errors happen and improves accuracy in scientific image analysis.
What this means in practice
- •For scientific software developers: Build scientific imaging tools that can explicitly separate task selection from execution for more reliable error diagnosis.
- •For automated quality control teams: Use transparent agent decision logs to distinguish between decision errors and execution failures in automated image inspections.
Authors
Shubhabrata Mukherjee, Jeffrey Donatelli, Daniela Ushizima
Abstract
Scientific image-analysis workflows require agents to select among diverse capabilities like description, detection, counting, and segmentation. Current multi-tool architectures conflate intent interpretation with visual execution, masking whether a failure arises from flawed reasoning (misrouting) or failed perception (invalid tool execution). This lack of auditability renders agents unreliable for high-stakes scientific automation. We introduce ZETA, an extensible architecture that operationalizes the decoupling of routing policy from action execution. ZETA resolves scientific intent before invoking specialized visual tools, treating measurement as a deterministic operation grounded in explicit geometry rather than generative output. This architecture exposes the "black box" of agent decision-making, allowing routing correctness and downstream execution to be evaluated as independent failure modes. To our knowledge, we provide the first controlled study in scientific vision-agent routing that isolates explicit routing-policy specification while holding the legal action interface fixed. Evaluation on a zero-shot benchmark of 51 scientific-intent queries shows that explicit routing semantics can substantially improve routing accuracy, particularly at semantically adjacent action boundaries, although improvements vary across models. For Qwen 3.8, explicit routing-policy specification improves segmentation routing accuracy from 6.2% to 75.0% and detection routing accuracy from 60.0% to 100%. ZETA's execution traces make routing decisions and downstream artifacts separately observable, providing a basis for distinguishing intent-selection errors from visual execution failures. ZETA establishes the necessary infrastructure to isolate execution errors from interpretive ones, a critical requirement for deploying reliable scientific vision agents.