IPDN: Image-enhanced Prompt Decoding Network for 3D Referring Expression Segmentation

TL;DR

This paper introduces IPDN, a novel network that enhances 3D referring expression segmentation by integrating multi-view images and task-driven prompts, effectively addressing feature and intent ambiguities.

Contribution

The paper proposes the Multi-view Semantic Embedding module and Prompt-Aware Decoder to improve reasoning in 3D-RES, outperforming existing methods.

Findings

IPDN outperforms state-of-the-art by 1.9 and 4.2 points in mIoU.

Multi-view semantic embedding improves spatial information retention.

Prompt-aware decoding enhances task-specific guidance.

Abstract

3D Referring Expression Segmentation (3D-RES) aims to segment point cloud scenes based on a given expression. However, existing 3D-RES approaches face two major challenges: feature ambiguity and intent ambiguity. Feature ambiguity arises from information loss or distortion during point cloud acquisition due to limitations such as lighting and viewpoint. Intent ambiguity refers to the model's equal treatment of all queries during the decoding process, lacking top-down task-specific guidance. In this paper, we introduce an Image enhanced Prompt Decoding Network (IPDN), which leverages multi-view images and task-driven information to enhance the model's reasoning capabilities. To address feature ambiguity, we propose the Multi-view Semantic Embedding (MSE) module, which injects multi-view 2D image information into the 3D scene and compensates for potential spatial information loss. To tackle intent ambiguity, we designed a Prompt-Aware Decoder (PAD) that guides the decoding process by deriving task-driven signals from the interaction between the expression and visual features. Comprehensive experiments demonstrate that IPDN outperforms the state-ofthe-art by 1.9 and 4.2 points in mIoU metrics on the 3D-RES and 3D-GRES tasks, respectively.