Single Tensor Cell Segmentation using Scalar Field Representations

2025-11-17

AI Summary

To address challenges in cell instance segmentation—including model complexity, high computational cost, and sensitivity to annotation noise—this paper proposes a lightweight and robust segmentation method based on scalar field representation. The method formulates a regularization-free robust regression framework: it generates semantically consistent, continuous scalar fields by solving the steady-state solutions of the Poisson and heat equations, inherently suppressing interference from erroneous annotations. Each cell instance is represented by a single tensor, and an end-to-end trainable architecture—integrating U-Net with PDE-residual minimization—is employed for learning. Efficient post-processing is achieved via watershed transformation. Evaluated on public benchmarks, the method achieves segmentation accuracy comparable to state-of-the-art approaches while significantly reducing training/inference time, memory footprint, and energy consumption—demonstrating strong practicality and deployment advantages for edge computing scenarios.

Technology Category

• Computer Vision: Segmentation
• Search and Optimization: Learning to Search
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments

Abstract

We investigate image segmentation of cells under the lens of scalar fields. Our goal is to learn a continuous scalar field on image domains such that its segmentation produces robust instances for cells present in images. This field is a function parameterized by the trained network, and its segmentation is realized by the watershed method. The fields we experiment with are solutions to the Poisson partial differential equation and a diffusion mimicking the steady-state solution of the heat equation. These solutions are obtained by minimizing just the field residuals, no regularization is needed, providing a robust regression capable of diminishing the adverse impacts of outliers in the training data and allowing for sharp cell boundaries. A single tensor is all that is needed to train a unet thus simplifying implementation, lowering training and inference times, hence reducing energy consumption, and requiring a small memory footprint, all attractive features in edge computing. We present competitive results on public datasets from the literature and show that our novel, simple yet geometrically insightful approach can achieve excellent cell segmentation results.

Problem

• Segmenting cells using continuous scalar field representations
• Achieving robust cell instance segmentation via watershed method
• Simplifying implementation with single tensor training for efficiency

Innovation

• Scalar field segmentation using watershed method
• Poisson and heat equation solutions for robust regression
• Single tensor UNet training for efficient edge computing