Filling the Pareto-Optimal Front for Affordance Segmentation on Embedded Devices Using RGB-D Cameras
arXiv:2607.28293 · doi:10.1109/JSEN.2025.3574506
Abstract
While depth sensors have the potential to complement RGB data for affordance segmentation in wearable robots, their usage seems to remain underexplored. The paper proposes two approaches: a reformulated version of hardware-aware neural architecture search, endowed with a newly designed search space to integrate depth (D) information into small-sized deep networks, and a dedicated fine-tuning approach, including a preprocessing layer to merge depth information with RGB data and make it compatible with conventional architectures. In both cases, those methods aim to generate solutions that benefit from modern (portable) hardware accelerators and overcome existing tiny-like approaches, which often fail to tackle critical scenarios due to the severe constraints set by the supporting hardware. Extensive experiments on a pair of real-world datasets demonstrate the effectiveness of the proposed method as compared with existing solutions. The approach presented in the paper generates, in most cases, solutions that identify the Pareto optimal front to balance generalization performance and hardware requirements. The paper also describes the supporting prototype, including a Jetson Nano board and a RealSense RGB-D camera. When considering the energy profile of the device, the overall system can attain real-time performances within an energy budget that is compatible with standard batteries, such as those used in smartphones.
References in corpus (9)
- Machine Learning for Microcontroller-Class Hardware: A Review
- The Deep Learning Compiler: A Comprehensive Survey
- Real-Time Radar-Based Gesture Detection and Recognition Built in an Edge-Computing Platform
- A Comprehensive Survey on Hardware-Aware Neural Architecture Search
- Enhancing Neural Architecture Search with Multiple Hardware Constraints for Deep Learning Model Deployment on Tiny IoT Devices
- Lightweight Neural Networks for Affordance Segmentation: Enhancement of the Decoder Module
- Are Dense Labels Always Necessary for 3D Object Detection from Point Cloud?
- Learning from 10 Demos: Generalisable and Sample-Efficient Policy Learning with Oriented Affordance Frames
- Affordance Labeling and Exploration: A Manifold-Based Approach