Automation of Waste Sorting with Deep Learning

By João Borges de Sousa, Ana Rebelo, Jaime S. Cardoso. 2019

71 citations recorded by the source corpus.

The importance of recycling is well known, either for environmental or economic reasons, it is impossible to escape it and the industry demands efficiency. Manual labour and traditional industrial sorting techniques are not capable of keeping up with the objectives demanded by the international community. Solutions based in computer vision techniques have the potential automate part of the waste handling tasks. In this paper, we propose a hierarchical deep learning approach for waste detection and classification in food trays. The proposed two-step approach retains the advantages of recent object detectors (as Faster R-CNN) and allows the classification task to be supported in higher resolution bounding boxes. Additionally, we also collect, annotate and make available to the scientific community a new dataset, named Labeled Waste in the Wild, for research and benchmark purposes. In the experimental comparison with standard deep learning approaches, the proposed hierarchical model shows better detection and classification performance.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition, Media Technology, Robotics & Control

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

1. Ping Hu
2. Kate Saenko
3. Berk Çalli

In Computer Vision and Pattern Recognition

1. Ping Hu
2. Berk Çalli
3. James Akl
4. Fadi M. Alladkani
5. Ziliang Zhu
6. Vitaly Ablavsky
7. Sarah Adel Bargal
8. Mohamed Abdelfattah

In Media Technology

1. Kate Saenko

In Robotics & Control

1. Kate Saenko
2. Berk Çalli

Citations in the corpus

• ZeroWaste Dataset: Towards Deformable Object Segmentation in Cluttered Scenes (2021, methodology): “…in the automated waste sorting space (e.g. [49, 50, 57]), they keep their data private, and the few existing open-source datasets [38, 48, 54, 60] are very limited in the amount of data and/or are generated in uncluttered environments, not representing the complexity of the domain (see…”
• Artificial intelligence for waste management in smart cities: a review (2023, unknown): “Moreover, Sousa et al. (2019) introduced a hierarchical deep learning method for sorting and identifying waste in food trays, which classifies waste into four categories based on material or ten categories based on shape using faster regions with convolutional neural network features.”