# Active learning with MaskAL reduces annotation effort for training Mask R-CNN

arXiv: https://arxiv.org/abs/2112.06586 · HTML: https://mlsanity.com/paper/2112.06586/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 15 of its 39 references that are in the map.
- First posted on arXiv 13 Dec 2021.

## Builds on (most cited first, 15 of 15)

- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Mask R-CNN](https://mlsanity.com/paper/1703.06870/) (2017)
- [Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning](https://mlsanity.com/paper/1506.02142/) (2015)
- [Aggregated Residual Transformations for Deep Neural Networks](https://mlsanity.com/paper/1611.05431/) (2016)
- [LVIS: A Dataset for Large Vocabulary Instance Segmentation](https://mlsanity.com/paper/1908.03195/) (2019)
- [A Survey of Deep Active Learning](https://mlsanity.com/paper/2009.00236/) (2020)
- [Concrete Dropout](https://mlsanity.com/paper/1705.07832/) (2017)
- [Deep Learning: A Critical Appraisal](https://mlsanity.com/paper/1801.00631/) (2018)
- [Dropout Sampling for Robust Object Detection in Open-Set Conditions](https://mlsanity.com/paper/1710.06677/) (2017)
- [The Limits and Potentials of Deep Learning for Robotics](https://mlsanity.com/paper/1804.06557/) (2018)
- [Active Learning for Deep Detection Neural Networks](https://mlsanity.com/paper/1911.09168/) (2019)
- [Deep Learning in Robotics: A Review of Recent Research](https://mlsanity.com/paper/1707.07217/) (2017)
- [Why AI is Harder Than We Think](https://mlsanity.com/paper/2104.12871/) (2021)
- [Active Learning with Point Supervision for Cost-Effective Panicle Detection in Cereal Crops](https://mlsanity.com/paper/1910.01789/) (2019)
- [Semi-supervised Active Learning for Instance Segmentation via Scoring Predictions](https://mlsanity.com/paper/2012.04829/) (2020)

## Built on by (most cited first, 7 of 7)

- [Label-Efficient Learning in Agriculture: A Comprehensive Review](https://mlsanity.com/paper/2305.14691/) (2023)
- [Active learning for efficient annotation in precision agriculture: a use-case on crop-weed semantic segmentation](https://mlsanity.com/paper/2404.02580/) (2024)
- [Interactively Learning Social Media Representations Improves News Source Factuality Detection](https://mlsanity.com/paper/2309.14966/) (2023)
- [Class-Incremental Learning of Plant and Disease Detection: Growing Branches with Knowledge Distillation](https://mlsanity.com/paper/2304.06619/) (2023)
- [Improved detection of discarded fish species through BoxAL active learning](https://mlsanity.com/paper/2410.04880/) (2024)
- [Adaptive path planning for efficient object search by UAVs in agricultural fields](https://mlsanity.com/paper/2504.02473/) (2025)
- [Development and evaluation of automated localisation and reconstruction of all fruits on tomato plants in a greenhouse based on multi-view perception and 3D multi-object tracking](https://mlsanity.com/paper/2211.02760/) (2022)

## Abstract (arXiv)

The generalisation performance of a convolutional neural network (CNN) is influenced by the quantity, quality, and variety of the training images. Training images must be annotated, and this is time consuming and expensive. The goal of our work was to reduce the number of annotated images needed to train a CNN while maintaining its performance. We hypothesised that the performance of a CNN can be improved faster by ensuring that the set of training images contains a large fraction of hard-to-classify images. The objective of our study was to test this hypothesis with an active learning method that can automatically select the hard-to-classify images. We developed an active learning method for Mask Region-based CNN (Mask R-CNN) and named this method MaskAL. MaskAL involved the iterative training of Mask R-CNN, after which the trained model was used to select a set of unlabelled images about which the model was most uncertain. The selected images were then annotated and used to retrain Mask R-CNN, and this was repeated for a number of sampling iterations. In our study, MaskAL was compared to a random sampling method on a broccoli dataset with five visually similar classes. MaskAL performed significantly better than the random sampling. In addition, MaskAL had the same performance after sampling 900 images as the random sampling had after 2300 images. Compared to a Mask R-CNN model that was trained on the entire training set (14,000 images), MaskAL achieved 93.9% of that model's performance with 17.9% of its training data. The random sampling achieved 81.9% of that model's performance with 16.4% of its training data. We conclude that by using MaskAL, the annotation effort can be reduced for training Mask R-CNN on a broccoli dataset with visually similar classes. Our software is available on https://github.com/pieterblok/maskal.