# Pouring Dynamics Estimation Using Gated Recurrent Units

arXiv: https://arxiv.org/abs/2105.12828 · HTML: https://mlsanity.com/paper/2105.12828/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 21 references that are in the map.
- First posted on arXiv 8 May 2021.

## Builds on (most cited first, 8 of 8)

- [Recurrent Neural Network Regularization](https://mlsanity.com/paper/1409.2329/) (2014)
- [Functional Object-Oriented Network for Manipulation Learning](https://mlsanity.com/paper/1902.01537/) (2019)
- [Functional Object-Oriented Network: Construction & Expansion](https://mlsanity.com/paper/1807.02189/) (2018)
- [Robot Gaining Accurate Pouring Skills through Self-Supervised Learning and Generalization](https://mlsanity.com/paper/2011.10150/) (2020)
- [Accurate Pouring with an Autonomous Robot Using an RGB-D Camera](https://mlsanity.com/paper/1810.03303/) (2018)
- [A Dataset of Daily Interactive Manipulation](https://mlsanity.com/paper/1807.00858/) (2018)
- [Manipulation Motion Taxonomy and Coding for Robots](https://mlsanity.com/paper/1910.00532/) (2019)
- [Learning to Pour](https://mlsanity.com/paper/1705.09021/) (2017)

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

- [The Sound of Water: Inferring Physical Properties from Pouring Liquids](https://mlsanity.com/paper/2411.11222/) (2024)

## Abstract (arXiv)

One of the most commonly performed manipulation in a human's daily life is pouring. Many factors have an effect on target accuracy, including pouring velocity, rotation angle, geometric of the source, and the receiving containers. This paper presents an approach to increase the repeatability and accuracy of the robotic manipulator by estimating the change in the amount of water of the pouring cup to a sequence of pouring actions using multiple layers of the deep recurrent neural network, especially gated recurrent units (GRU). The proposed GRU model achieved a validation mean squared error as low as 1e-4 (lbf) for the predicted value of weight f(t). This paper contains a comprehensive evaluation and analysis of numerous experiments with various designs of recurrent neural networks and hyperparameters fine-tuning.