Parameter-Free Adaptive Multi-Scale Channel-Spatial Attention Aggregation framework for 3D Indoor Semantic Scene Completion Toward Assisting Visually Impaired

2026-02-18

AI Summary

This work addresses the structural instability in existing monocular 3D semantic scene completion methods, which stems from the lack of explicit modeling of voxel feature reliability and ineffective cross-scale information regulation, leading to projection diffusion and feature entanglement. Building upon the MonoScene framework, we propose a parameter-free adaptive multi-scale channel-spatial parallel attention mechanism to calibrate voxel feature reliability, along with a hierarchical adaptive gating strategy to stabilize multi-scale feature fusion between the encoder and decoder. Evaluated on the NYUv2 benchmark, our approach achieves an SSC mIoU of 27.25% (+0.31) and an SC IoU of 43.10% (+0.59), while remaining efficiently deployable on NVIDIA Jetson embedded platforms.

Technology Category

• Computer Vision: Scene Analysis & Understanding
• Machine Learning: Multi-instance/Multi-view Learning
• Knowledge Representation and Reasoning: Nonmonotonic Reasoning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Security and Privacy: Large-scale security measurements

Abstract

In indoor assistive perception for visually impaired users, 3D Semantic Scene Completion (SSC) is expected to provide structurally coherent and semantically consistent occupancy under strictly monocular vision for safety-critical scene understanding. However, existing monocular SSC approaches often lack explicit modeling of voxel-feature reliability and regulated cross-scale information propagation during 2D-3D projection and multi-scale fusion, making them vulnerable to projection diffusion and feature entanglement and thus limiting structural stability.To address these challenges, this paper presents an Adaptive Multi-scale Attention Aggregation (AMAA) framework built upon the MonoScene pipeline. Rather than introducing a heavier backbone, AMAA focuses on reliability-oriented feature regulation within a monocular SSC framework. Specifically, lifted voxel features are jointly calibrated in semantic and spatial dimensions through parallel channel-spatial attention aggregation, while multi-scale encoder-decoder fusion is stabilized via a hierarchical adaptive feature-gating strategy that regulates information injection across scales.Experiments on the NYUv2 benchmark demonstrate consistent improvements over MonoScene without significantly increasing system complexity: AMAA achieves 27.25% SSC mIoU (+0.31) and 43.10% SC IoU (+0.59). In addition, system-level deployment on an NVIDIA Jetson platform verifies that the complete AMAA framework can be executed stably on embedded hardware. Overall, AMAA improves monocular SSC quality and provides a reliable and deployable perception framework for indoor assistive systems targeting visually impaired users.

Problem

• Semantic Scene Completion
• Monocular Vision
• Visually Impaired Assistance
• 3D Scene Understanding
• Feature Reliability

Innovation

• Parameter-Free
• Channel-Spatial Attention
• Multi-scale Fusion
• Semantic Scene Completion
• Monocular 3D Perception