Abstract:In the field of referring segmentation, top-down methods and bottom-up methods are the two prevailing approaches. Both of these methods inevitably exhibit certain drawbacks. Top-down methods are susceptible to Polar Negative (PN) errors due to their limited understanding of multi-modal fine-grained features. Bottom-up methods lack macro-level object positional information, making them susceptible to Inferior Positive (IP) errors. However, we find that the two approaches are highly complementary in addressing their respective weaknesses, but combining them directly through a simple average does not yield complementary advantages. Therefore, we proposed a hierarchical collaboration approach to explore the complementary characteristics of the existing two methods from the perspectives of fusion and interaction, aiming to achieve more precise segmentation results. We proposed the Complementary Feature Interaction (CFI) module, which enables top-down methods to access fine-grained information and allows bottom-up approaches to obtain object positional information interactively. Regarding integration, Gaussian Scoring Integration (GSI) models the Gaussian performance distributions of two branches and performs weighted integration by sampling confidence scores from these distributions. We integrate various top-down and bottom-up methods within the proposed architecture and conduct experiments on three standard datasets. The experimental results demonstrate that our method outperforms the state-of-the-art independent segmentation algorithms. On the RefCOCO validation, test A and test B datasets, our proposed method achieved IoU scores of 77.51, 79.12, and 72.79, respectively. Extensive experiments demonstrate that our method can significantly improve segmentation accuracy when fusing different sub-methods.

Repairing bad co-segmentation using its quality evaluation and segment propagation.

A Novel Quality Evaluating Method for Over-Segmentation Approaches Using Real-Time Boundary Information

Interactive Color Image Segmentation by Region Growing Combined with Image Enhancement Based on Bezier Model

Interactive shape co-segmentation via label propagation

Improving Image Segmentation with Boundary Patch Refinement

Coarse-to-fine Image Co-segmentation with Intra and Inter Rank Constraints

Evaluation of Segmentation Quality via Adaptive Composition of Reference Segmentations.

Noise-aware co-segmentation with local and global priors.

Seeds-Based Part Segmentation by Seeds Propagation and Region Convexity Decomposition.

Image Co-segmentation by Incorporating Color Reward Strategy and Active Contours Model

Weak Supervision Learning for Object Co-Segmentation

Improvement of Region-Merging Image Segmentation Accuracy Using Multiple Merging Criteria

Object Cosegmentation by Similarity Propagation with Saliency Information and Objectness Frequency Map

Cosegmentation from Similar Backgrounds

Segmentation Quality Evaluation Based on Multi-Scale Convolutional Neural Networks

Segmentation Quality Evaluation Using Region-Based Precision and Recall Measures for Remote Sensing Images

Hierarchical Collaboration for Referring Image Segmentation

Part Propagation For Local Part Segmentation

CycleSegNet: Object Co-segmentation with Cycle Refinement and Region Correspondence

Another Look on Region Merging Procedure from Seed Region Shift for High-Resolution Remote Sensing Image Segmentation

Image Co-Segmentation Via Active Contours