# Virtual Temporal Samples for Recurrent Neural Networks: applied to semantic segmentation in agriculture

arXiv: https://arxiv.org/abs/2106.10118 · HTML: https://mlsanity.com/paper/2106.10118/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 14 of its 34 references that are in the map.
- First posted on arXiv 18 Jun 2021.

## Builds on (most cited first, 14 of 14)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [Fully Convolutional Networks for Semantic Segmentation](https://mlsanity.com/paper/1411.4038/) (2014)
- [The Cityscapes Dataset for Semantic Urban Scene Understanding](https://mlsanity.com/paper/1604.01685/) (2016)
- [Learning Deconvolution Network for Semantic Segmentation](https://mlsanity.com/paper/1505.04366/) (2015)
- [Deep learning in agriculture: A survey](https://mlsanity.com/paper/1807.11809/) (2018)
- [Spatial-Temporal Recurrent Neural Network for Emotion Recognition](https://mlsanity.com/paper/1705.04515/) (2017)
- [PATHoBot: A Robot for Glasshouse Crop Phenotyping and Intervention](https://mlsanity.com/paper/2010.16272/) (2020)
- [Fully Convolutional Networks with Sequential Information for Robust Crop and Weed Detection in Precision Farming](https://mlsanity.com/paper/1806.03412/) (2018)
- [Semantic Segmentation of Video Sequences with Convolutional LSTMs](https://mlsanity.com/paper/1905.01058/) (2019)
- [A two-stage 3D Unet framework for multi-class segmentation on full resolution image](https://mlsanity.com/paper/1804.04341/) (2018)
- [Weakly Supervised Image Annotation and Segmentation with Objects and Attributes](https://mlsanity.com/paper/1708.02459/) (2017)
- [Feedback Attention for Cell Image Segmentation](https://mlsanity.com/paper/2008.06474/) (2020)

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

- [GrowliFlower: An image time series dataset for GROWth analysis of cauLIFLOWER](https://mlsanity.com/paper/2204.00294/) (2022)
- [Explicitly incorporating spatial information to recurrent networks for agriculture](https://mlsanity.com/paper/2206.13406/) (2022)
- [Zero-shot Hierarchical Plant Segmentation via Foundation Segmentation Models and Text-to-image Attention](https://mlsanity.com/paper/2509.09116/) (2025)
- [Panoptic One-Click Segmentation: Applied to Agricultural Data](https://mlsanity.com/paper/2303.08689/) (2023)
- [BonnBot-I Plus: A Bio-diversity Aware Precise Weed Management Robotic Platform](https://mlsanity.com/paper/2405.09118/) (2024)
- [AgriField-40K: Adapting Vision Models to Agriculture With Efficient Continual Pretraining](https://mlsanity.com/paper/2608.07984/) (2026)
- [DropClick: Semi-Automated One-Click Segmentation for Agricultural Robotic Data](https://mlsanity.com/paper/2609.03680/) (2026)

## Abstract (arXiv)

This paper explores the potential for performing temporal semantic segmentation in the context of agricultural robotics without temporally labelled data. We achieve this by proposing to generate virtual temporal samples from labelled still images. By exploiting the relatively static scene and assuming that the robot (camera) moves we are able to generate virtually labelled temporal sequences with no extra annotation effort. Normally, to train a recurrent neural network (RNN), labelled samples from a video (temporal) sequence are required which is laborious and has stymied work in this direction. By generating virtual temporal samples, we demonstrate that it is possible to train a lightweight RNN to perform semantic segmentation on two challenging agricultural datasets. Our results show that by training a temporal semantic segmenter using virtual samples we can increase the performance by an absolute amount of 4.6 and 4.9 on sweet pepper and sugar beet datasets, respectively. This indicates that our virtual data augmentation technique is able to accurately classify agricultural images temporally without the use of complicated synthetic data generation techniques nor with the overhead of labelling large amounts of temporal sequences.