Fully Aligned Network for Referring Image Segmentation
This work addresses the challenge of cross-modal comprehension in RIS for computer vision applications, representing an incremental improvement over existing attention-based methods.
The paper tackles the problem of achieving fine-grained alignment between language and vision in Referring Image Segmentation (RIS) by proposing a Fully Aligned Network (FAN) that follows explicit interaction principles, resulting in state-of-the-art performance on benchmarks like RefCOCO, RefCOCO+, and G-Ref.
This paper focuses on the Referring Image Segmentation (RIS) task, which aims to segment objects from an image based on a given language description. The critical problem of RIS is achieving fine-grained alignment between different modalities to recognize and segment the target object. Recent advances using the attention mechanism for cross-modal interaction have achieved excellent progress. However, current methods tend to lack explicit principles of interaction design as guidelines, leading to inadequate cross-modal comprehension. Additionally, most previous works use a single-modal mask decoder for prediction, losing the advantage of full cross-modal alignment. To address these challenges, we present a Fully Aligned Network (FAN) that follows four cross-modal interaction principles. Under the guidance of reasonable rules, our FAN achieves state-of-the-art performance on the prevalent RIS benchmarks (RefCOCO, RefCOCO+, G-Ref) with a simple architecture.