TrashDet: Iterative Neural Architecture Search for Efficient Waste Detection

TL;DR

This paper introduces TrashDet, a neural architecture search framework for creating efficient waste detectors suitable for TinyML devices, achieving high accuracy and low energy consumption on resource-constrained hardware.

Contribution

It presents a hardware-aware iterative neural architecture search method that generates scalable, deployment-ready trash detectors optimized for TinyML constraints.

Findings

TrashDet-l achieves 19.5 mAP50 with 30.5M parameters, outperforming prior detectors.

The TrashDet family offers scalable models from 1.2M to 30.5M parameters with varying accuracy.

On MAX78002 microcontroller, TrashDet variants significantly reduce energy, latency, and power consumption.

Abstract

This paper addresses trash detection on the TACO dataset under strict TinyML constraints using an iterative hardware-aware neural architecture search framework targeting edge and IoT devices. The proposed method constructs a Once-for-All-style ResDets supernet and performs iterative evolutionary search that alternates between backbone and neck/head optimization, supported by a population passthrough mechanism and an accuracy predictor to reduce search cost and improve stability. This framework yields a family of deployment-ready detectors, termed TrashDets. On a five-class TACO subset (paper, plastic, bottle, can, cigarette), the strongest variant, TrashDet-l, achieves 19.5 mAP50 with 30.5M parameters, improving accuracy by up to 3.6 mAP50 over prior detectors while using substantially fewer parameters. The TrashDet family spans 1.2M to 30.5M parameters with mAP50 values between 11.4 and 19.5, providing scalable detector options for diverse TinyML deployment budgets on resource-constrained hardware. On the MAX78002 microcontroller with the TrashNet dataset, two specialized variants, TrashDet-ResNet and TrashDet-MBNet, jointly dominate the ai87-fpndetector baseline, with TrashDet-ResNet achieving 7525~$\mu$J energy per inference at 26.7 ms latency and 37.45 FPS, and TrashDet-MBNet improving mAP50 by 10.2%; together they reduce energy consumption by up to 88%, latency by up to 78%, and average power by up to 53% compared to existing TinyML detectors.