# A Multi-Level Approach to Waste Object Segmentation

arXiv: https://arxiv.org/abs/2007.04259 · HTML: https://mlsanity.com/paper/2007.04259/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 Jul 2020.

## Built on by (most cited first, 11 of 11)

- [Waste detection in Pomerania: non-profit project for detecting waste in environment](https://mlsanity.com/paper/2105.06808/) (2021)
- [Material Measurement Units for a Circular Economy: Foundations through a Review](https://mlsanity.com/paper/2103.01997/) (2021)
- [GlobalWasteData: A Large-Scale, Integrated Dataset for Robust Waste Classification and Environmental Monitoring](https://mlsanity.com/paper/2602.07463/) (2026)
- [Temporal-consistent CAMs for Weakly Supervised Video Segmentation in Waste Sorting](https://mlsanity.com/paper/2502.01455/) (2025)
- [Learning Using Privileged Information for Litter Detection](https://mlsanity.com/paper/2508.04124/) (2025)
- [StreetView-Waste: A Multi-Task Dataset for Urban Waste Management](https://mlsanity.com/paper/2511.16440/) (2025)
- [An Ensemble Learning Approach towards Waste Segmentation in Cluttered Environment](https://mlsanity.com/paper/2602.13681/) (2026)
- [Garbage Segmentation and Attribute Analysis by Robotic Dogs](https://mlsanity.com/paper/2404.18112/) (2024)
- [AI for Green Spaces: Leveraging Autonomous Navigation and Computer Vision for Park Litter Removal](https://mlsanity.com/paper/2601.11876/) (2026)
- [The Garbage Dataset (GD): A Multi-Class Image Benchmark for Automated Waste Segregation](https://mlsanity.com/paper/2602.10500/) (2026)
- [MDWD: A Street-Level Dataset for Municipal Solid Waste Detection in Dense Urban Environments](https://mlsanity.com/paper/2608.00257/) (2026)

## Abstract (arXiv)

We address the problem of localizing waste objects from a color image and an optional depth image, which is a key perception component for robotic interaction with such objects. Specifically, our method integrates the intensity and depth information at multiple levels of spatial granularity. Firstly, a scene-level deep network produces an initial coarse segmentation, based on which we select a few potential object regions to zoom in and perform fine segmentation. The results of the above steps are further integrated into a densely connected conditional random field that learns to respect the appearance, depth, and spatial affinities with pixel-level accuracy. In addition, we create a new RGBD waste object segmentation dataset, MJU-Waste, that is made public to facilitate future research in this area. The efficacy of our method is validated on both MJU-Waste and the Trash Annotation in Context (TACO) dataset.