Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

DWaste: Greener AI for Waste Sorting using Mobile and Edge Devices

Suman Kunwar

TL;DR

This work addresses the challenge of real-time waste sorting on resource-constrained devices by benchmark-ing classification and object-detection models on a seven-category recycling dataset. It introduces a CBAM-enhanced YOLO variant and leverages quantization to drastically reduce memory and energy use, enabling edge and mobile deployment. The results highlight a trade-off between accuracy and efficiency, with lightweight detectors offering competitive mAP and far lower power draw, while quantization amplifies these gains. Ultimately, YOLOv8n-CBAM is deployed in the DWaste app and on an edge device, illustrating a practical path toward greener AI in urban waste management.

Abstract

The rise of convenience packaging has led to generation of enormous waste, making efficient waste sorting crucial for sustainable waste management. To address this, we developed DWaste, a computer vision-powered platform designed for real-time waste sorting on resource-constrained smartphones and edge devices, including offline functionality. We benchmarked various image classification models (EfficientNetV2S/M, ResNet50/101, MobileNet) and object detection (YOLOv8n, YOLOv11n) including our purposed YOLOv8n-CBAM model using our annotated dataset designed for recycling. We found a clear trade-off between accuracy and resource consumption: the best classifier, EfficientNetV2S, achieved high accuracy(~ 96%) but suffered from high latency (~ 0.22s) and elevated carbon emissions. In contrast, lightweight object detection models delivered strong performance (up to 80% mAP) with ultra-fast inference (~ 0.03s) and significantly smaller model sizes (< 7MB ), making them ideal for real-time, low-power use. Model quantization further maximized efficiency, substantially reducing model size and VRAM usage by up to 75%. Our work demonstrates the successful implementation of "Greener AI" models to support real-time, sustainable waste sorting on edge devices.

DWaste: Greener AI for Waste Sorting using Mobile and Edge Devices

TL;DR

This work addresses the challenge of real-time waste sorting on resource-constrained devices by benchmark-ing classification and object-detection models on a seven-category recycling dataset. It introduces a CBAM-enhanced YOLO variant and leverages quantization to drastically reduce memory and energy use, enabling edge and mobile deployment. The results highlight a trade-off between accuracy and efficiency, with lightweight detectors offering competitive mAP and far lower power draw, while quantization amplifies these gains. Ultimately, YOLOv8n-CBAM is deployed in the DWaste app and on an edge device, illustrating a practical path toward greener AI in urban waste management.

Abstract

The rise of convenience packaging has led to generation of enormous waste, making efficient waste sorting crucial for sustainable waste management. To address this, we developed DWaste, a computer vision-powered platform designed for real-time waste sorting on resource-constrained smartphones and edge devices, including offline functionality. We benchmarked various image classification models (EfficientNetV2S/M, ResNet50/101, MobileNet) and object detection (YOLOv8n, YOLOv11n) including our purposed YOLOv8n-CBAM model using our annotated dataset designed for recycling. We found a clear trade-off between accuracy and resource consumption: the best classifier, EfficientNetV2S, achieved high accuracy(~ 96%) but suffered from high latency (~ 0.22s) and elevated carbon emissions. In contrast, lightweight object detection models delivered strong performance (up to 80% mAP) with ultra-fast inference (~ 0.03s) and significantly smaller model sizes (< 7MB ), making them ideal for real-time, low-power use. Model quantization further maximized efficiency, substantially reducing model size and VRAM usage by up to 75%. Our work demonstrates the successful implementation of "Greener AI" models to support real-time, sustainable waste sorting on edge devices.
Paper Structure (6 sections, 8 figures, 1 table)

This paper contains 6 sections, 8 figures, 1 table.

Table of Contents

  1. Introduction
  2. Materials and Methods
  3. Dataset and Preprocessing
  4. Model Training
  5. Results and Discussion
  6. Conclusion

Figures (8)

  • Figure 1: Dataset class distribution
  • Figure 2: Sample annotated images from our dataset
  • Figure 3: Architecture of the proposed Improved YOLOv8n model with CBAM-enhanced backbone
  • Figure 4: Carbon emission by stage for each model
  • Figure 5: Classification models VRAM usage under inference
  • ...and 3 more figures