Deep ContourFlow: Advancing Active Contours with Deep Learning

TL;DR

Deep ContourFlow (DCF) addresses histology segmentation with scarce labels by uniting unsupervised active contours and one-shot learning using multi-scale CNN features from a pre-trained network. It introduces two differentiable contour mappings, F_cm and F_cd, to drive contour evolution without full supervision and extends it to one-shot segmentation via isolines and distance maps, plus a cosine-based classification step. On the AIDPATH kidney dataset, DCF achieves competitive Dice and related metrics, outperforming several one-shot baselines while highlighting robustness to initialization and complex textures in histology. The approach reduces labeling requirements and provides an open-source implementation, offering a practical tool for precise histology segmentation with minimal supervision.

Abstract

This paper introduces a novel approach that combines unsupervised active contour models with deep learning for robust and adaptive image segmentation. Indeed, traditional active contours, provide a flexible framework for contour evolution and learning offers the capacity to learn intricate features and patterns directly from raw data. Our proposed methodology leverages the strengths of both paradigms, presenting a framework for both unsupervised and one-shot approaches for image segmentation. It is capable of capturing complex object boundaries without the need for extensive labeled training data. This is particularly required in histology, a field facing a significant shortage of annotations due to the challenging and time-consuming nature of the annotation process. We illustrate and compare our results to state of the art methods on a histology dataset and show significant improvements.

Figures (12)

• Figure 1: Example of a polygon and representation of the oriented angles connecting an interior point $x$ to its nodes $C_j$ along the contour.
• Figure 2: Approximation of the mask of a circle using $F_{cm}$ (top row) and the distance map using $F_{cd}$ (bottom row) when increasing k with $n_{nodes}=100$.
• Figure 3: Approximation of the mask of a circle using $F_{cm}$ (middle row) and the distance map using $F_{cd}$ (bottom row) when increasing $n_{nodes}$ with $k=10^{5}$.
• Figure 4: Illustration of the effects of the operators used in the iteration adjustments. From top to bottom and left to right: Clean, Clip, Interp and Blur operators are used respectively to delete loops, impose a maximum displacement, resample the points along the contour and regularize the displacement.
• Figure 5: Unsupervised DCF: evolution of the contour on four real-life images when varying the initial contour $C_0$.