# Neural Inverse Kinematics

arXiv: https://arxiv.org/abs/2205.10837 · HTML: https://mlsanity.com/paper/2205.10837/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 7 of its 32 references that are in the map.
- First posted on arXiv 22 May 2022.

## Builds on (most cited first, 7 of 7)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [On the Properties of Neural Machine Translation: Encoder-Decoder Approaches](https://mlsanity.com/paper/1409.1259/) (2014)
- [From Softmax to Sparsemax: A Sparse Model of Attention and Multi-Label Classification](https://mlsanity.com/paper/1602.02068/) (2016)
- [Active Learning of Inverse Models with Intrinsically Motivated Goal Exploration in Robots](https://mlsanity.com/paper/1301.4862/) (2013)
- [Encoding Physical Constraints in Differentiable Newton-Euler Algorithm](https://mlsanity.com/paper/2001.08861/) (2020)
- [Meta Internal Learning](https://mlsanity.com/paper/2110.02900/) (2021)
- [Computing Robust Inverse Kinematics Under Uncertainty](https://mlsanity.com/paper/1910.10852/) (2019)

## Built on by (most cited first, 4 of 4)

- [SparsePoser: Real-time Full-body Motion Reconstruction from Sparse Data](https://mlsanity.com/paper/2311.02191/) (2023)
- [SMPL-IK: Learned Morphology-Aware Inverse Kinematics for AI Driven Artistic Workflows](https://mlsanity.com/paper/2208.08274/) (2022)
- [A Design Co-Pilot for Task-Tailored Manipulators](https://mlsanity.com/paper/2509.13077/) (2025)
- [Amortized Inverse Kinematics via Graph Attention for Real-Time Human Avatar Animation](https://mlsanity.com/paper/2604.16629/) (2026)

## Abstract (arXiv)

Inverse kinematic (IK) methods recover the parameters of the joints, given the desired position of selected elements in the kinematic chain. While the problem is well-defined and low-dimensional, it has to be solved rapidly, accounting for multiple possible solutions. In this work, we propose a neural IK method that employs the hierarchical structure of the problem to sequentially sample valid joint angles conditioned on the desired position and on the preceding joints along the chain. In our solution, a hypernetwork f recovers the parameters of multiple primary networks g1,g2,dots,gN, where N is the number of joints, such that each gi outputs a distribution of possible joint angles, and is conditioned on the sampled values obtained from the previous primary networks gj, j<i. The hypernetwork can be trained on readily available pairs of matching joint angles and positions, without observing multiple solutions. At test time, a high-variance joint distribution is presented, by sampling sequentially from the primary networks. We demonstrate the advantage of the proposed method both in comparison to other IK methods for isolated instances of IK and with regard to following the path of the end effector in Cartesian space.