An official website of the United States government
A central challenge in machine learning deployment is maintaining accurate and updated models as the deployment environment changes over time. We present a hardware/software framework for simultaneous training and inference for monocular depth estimation on edge devices. Our proposed framework can be used as a hardware/software co-design tool that enables continual and online federated learning on edge devices. Our results show real-time training and inference performance, demonstrating the feasibility of online learning on edge devices.
more »
« less
This content will become publicly available on May 6, 2026
Demo Abstract: Lightweight Training and Inference for Self-Supervised Depth Estimation on Edge Devices
Not AvailablDeploying monocular depth estimation on resource-constrained edge devices is a significant challenge, particularly when attempting to perform both training and inference concurrently. Current lightweight, self-supervised approaches typically rely on complex frameworks that are hard to implement and deploy in real-world settings. To address this gap, we introduce the first framework for Lightweight Training and Inference (LITI) that combines ready-to-deploy models with streamlined code and fully functional, parallel training and inference pipelines. Our experiments show various models being deployed for inference, training, or both inference and training, leveraging inputs from a real-time RGB camera sensor. Thus, our framework enables training and inference on resource-constrained edge devices for complex applications such as depth estimation.
more »
« less
- Award ID(s):
- 2107085
- PAR ID:
- 10639030
- Publisher / Repository:
- ACM
- Date Published:
- Page Range / eLocation ID:
- 714 to 715
- Subject(s) / Keyword(s):
- Edge Devices Energy Efficiency Monocular Depth Estimation Self-Supervised Learning
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Modern heavy vehicles rely on insecure protocols (CAN and SAE-J1939) to facilitate communication between the embedded devices that control their various subsys- tems. Due to the growing integration of wireless-enabled embedded devices, vehicles are becoming increasingly vulnerable to remote cyberattacks against their embedded networks. We propose an efficient deep-learning-based approach for mitigating such attacks through real-time J1939 signal reconstruction. Our approach uses random feature masking during training to build a generalized model of a vehicle’s network. To reduce the computa- tional and storage burden of the model, we employ 8-bit Quantization-Aware Training (QAT), enabling its deploy- ment on resource-constrained embedded devices while maintaining high performance. We evaluate Transformer and LSTM-based architectures, demonstrating that both effectively reconstruct signals with minimal computa- tional and storage overhead. Our approach achieves sig- nal reconstruction with error levels below 1% of their operating range while maintaining a very low storage footprint of under 1 MB, demonstrating that lightweight deep-learning models can enhance resiliency against real- time attacks in heavy vehicles.more » « less
-
Word-aware sentiment analysis has posed a significant challenge over the past decade. Despite the considerable efforts of recent language models, achieving a lightweight representation suitable for deployment on resource-constrained edge devices remains a crucial concern. This study proposes a novel solution by merging two emerging paradigms, the Word2Vec language model and Hyperdimensional Computing, and introduces an innovative framework named Word2HyperVec. Our framework prioritizes model size and facilitates low-power processing during inference by incorporating embeddings into a binary space. Our solution demonstrates significant advantages, consuming only 2.2 W, up to 1.81 × more efficient than alternative learning models such as support vector machines, random forest, and multi-layer perceptron.more » « less
-
Word-aware sentiment analysis has posed a significant challenge over the past decade. Despite the considerable efforts of recent language models, achieving a lightweight representation suitable for deployment on resource-constrained edge devices remains a crucial concern. This study proposes a novel solution by merging two emerging paradigms, the Word2Vec language model and Hyperdimensional Computing, and introduces an innovative framework named Word2HyperVec. Our framework prioritizes model size and facilitates low-power processing during inference by incorporating embeddings into a binary space. Our solution demonstrates significant advantages, consuming only 2.2 W, up to 1.81 × more efficient than alternative learning models such as support vector machines, random forest, and multi-layer perceptron.more » « less
-
Classification tasks on ultra-lightweight devices demand devices that are resource-constrained and deliver swift responses. Binary Vector Symbolic Architecture (VSA) is a promising approach due to its minimal memory requirements and fast execution times compared to traditional machine learning (ML) methods. Nonetheless, binary VSA's practicality is limited by its inferior inference performance and a design that prioritizes algorithmic over hardware optimization. This paper introduces UniVSA, a co-optimized binary VSA framework for both algorithm and hardware. UniVSA not only significantly enhances inference accuracy beyond current state-of-the-art binary VSA models but also reduces memory footprints. It incorporates novel, lightweight modules and design flow tailored for optimal hardware performance. Experimental results show that UniVSA surpasses traditional ML methods in terms of performance on resource-limited devices, achieving smaller memory usage, lower latency, reduced resource demand, and decreased power consumption.more » « less
