# On Domain-Specific Pre-Training for Effective Semantic Perception in Agricultural Robotics

arXiv: https://arxiv.org/abs/2303.12499 · HTML: https://mlsanity.com/paper/2303.12499/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 49 references that are in the map.
- First posted on arXiv 22 Mar 2023.

## Builds on (most cited first, 21 of 21)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [Momentum Contrast for Unsupervised Visual Representation Learning](https://mlsanity.com/paper/1911.05722/) (2019)
- [mixup: Beyond Empirical Risk Minimization](https://mlsanity.com/paper/1710.09412/) (2017)
- [Fully Convolutional Networks for Semantic Segmentation](https://mlsanity.com/paper/1411.4038/) (2014)
- [Bootstrap your own latent: A new approach to self-supervised Learning](https://mlsanity.com/paper/2006.07733/) (2020)
- [Unsupervised Learning of Visual Features by Contrasting Cluster Assignments](https://mlsanity.com/paper/2006.09882/) (2020)
- [Improved Baselines with Momentum Contrastive Learning](https://mlsanity.com/paper/2003.04297/) (2020)
- [Barlow Twins: Self-Supervised Learning via Redundancy Reduction](https://mlsanity.com/paper/2103.03230/) (2021)
- [Random Erasing Data Augmentation](https://mlsanity.com/paper/1708.04896/) (2017)
- [AutoAugment: Learning Augmentation Policies from Data](https://mlsanity.com/paper/1805.09501/) (2018)
- [Rethinking ImageNet Pre-training](https://mlsanity.com/paper/1811.08883/) (2018)
- [CNN Features off-the-shelf: an Astounding Baseline for Recognition](https://mlsanity.com/paper/1403.6382/) (2014)
- [Coupled Generative Adversarial Networks](https://mlsanity.com/paper/1606.07536/) (2016)
- [Learning to Compose Domain-Specific Transformations for Data Augmentation](https://mlsanity.com/paper/1709.01643/) (2017)
- [Real-time Semantic Segmentation of Crop and Weed for Precision Agriculture Robots Leveraging Background Knowledge in CNNs](https://mlsanity.com/paper/1709.06764/) (2017)
- [Pixel-Level Domain Transfer](https://mlsanity.com/paper/1603.07442/) (2016)
- [Fully Convolutional Networks with Sequential Information for Robust Crop and Weed Detection in Precision Farming](https://mlsanity.com/paper/1806.03412/) (2018)
- [Building an Aerial-Ground Robotics System for Precision Farming: An Adaptable Solution](https://mlsanity.com/paper/1911.03098/) (2019)
- [Joint Stem Detection and Crop-Weed Classification for Plant-specific Treatment in Precision Farming](https://mlsanity.com/paper/1806.03413/) (2018)
- [Unsupervised Domain Adaptation For Plant Organ Counting](https://mlsanity.com/paper/2009.01081/) (2020)

## Built on by (most cited first, 1 of 1)

- [INoD: Injected Noise Discriminator for Self-Supervised Representation Learning in Agricultural Fields](https://mlsanity.com/paper/2303.18101/) (2023)

## Abstract (arXiv)

Agricultural robots have the prospect to enable more efficient and sustainable agricultural production of food, feed, and fiber. Perception of crops and weeds is a central component of agricultural robots that aim to monitor fields and assess the plants as well as their growth stage in an automatic manner. Semantic perception mostly relies on deep learning using supervised approaches, which require time and qualified workers to label fairly large amounts of data. In this paper, we look into the problem of reducing the amount of labels without compromising the final segmentation performance. For robots operating in the field, pre-training networks in a supervised way is already a popular method to reduce the number of required labeled images. We investigate the possibility of pre-training in a self-supervised fashion using data from the target domain. To better exploit this data, we propose a set of domain-specific augmentation strategies. We evaluate our pre-training on semantic segmentation and leaf instance segmentation, two important tasks in our domain. The experimental results suggest that pre-training with domain-specific data paired with our data augmentation strategy leads to superior performance compared to commonly used pre-trainings. Furthermore, the pre-trained networks obtain similar performance to the fully supervised with less labeled data.