CVEvolve: Autonomous Algorithm Discovery for Unstructured Scientific Data Processing

TL;DR

CVEvolve is an autonomous, zero-code system that leverages LLMs to discover and improve algorithms for processing complex scientific images, aiding domain scientists without extensive technical expertise.

Contribution

It introduces CVEvolve, a novel autonomous agentic framework combining search strategies and LLMs for scientific data algorithm discovery without coding.

Findings

CVEvolve outperforms baseline methods in image registration, peak detection, and segmentation tasks.

Holdout testing helps identify algorithms with better generalization.

The system enables domain scientists to develop practical algorithms from unstructured data.

Abstract

Scientific data processing often requires task-specific algorithms or AI models, creating a barrier for domain scientists who need to analyze their data but may not have extensive computing or image-processing expertise. This barrier is especially pronounced when data are noisy, have a high dynamic range, are sparsely labeled, or are only loosely specified. We introduce CVEvolve, an autonomous agentic harness with a zero-code interface for scientific data-processing algorithm discovery. CVEvolve combines a multi-round search strategy with tools for code execution, evaluation implementation, history management, holdout testing, and optional inspection of scientific data and visual outputs. The search alternates between discovery and improvement actions, and uses lineage-aware stochastic candidate sampling to balance exploration and exploitation. We demonstrate CVEvolve on x-ray fluorescence microscopy image registration, Bragg peak detection, and high-energy diffraction microscopy image segmentation. Across these tasks, CVEvolve discovers algorithms that improve over baseline methods, while holdout test tracking helps identify candidates that generalize better than later over-optimized alternatives. These results show that zero-code, autonomous LLM-powered algorithm development can help domain scientists turn unstructured scientific image data into practical algorithms and downstream scientific discoveries.