# Human Activity Recognition Using 3D Orthogonally-projected EfficientNet on Radar Time-Range-Doppler Signature

arXiv: https://arxiv.org/abs/2111.12375 · HTML: https://mlsanity.com/paper/2111.12375/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 4 of its 23 references that are in the map.
- First posted on arXiv 24 Nov 2021.

## Builds on (most cited first, 4 of 4)

- [EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks](https://mlsanity.com/paper/1905.11946/) (2019)
- [Non-local Neural Networks](https://mlsanity.com/paper/1711.07971/) (2017)
- [Can Spatiotemporal 3D CNNs Retrace the History of 2D CNNs and ImageNet?](https://mlsanity.com/paper/1711.09577/) (2017)
- [Video-based Person Re-identification via 3D Convolutional Networks and Non-local Attention](https://mlsanity.com/paper/1807.05073/) (2018)

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

- [Deep learning for radar data exploitation of autonomous vehicle](https://mlsanity.com/paper/2203.08038/) (2022)

## Abstract (arXiv)

In radar activity recognition, 2D signal representations such as spectrogram, cepstrum and cadence velocity diagram are often utilized, while range information is often neglected. In this work, we propose to utilize the 3D time-range-Doppler (TRD) representation, and design a 3D Orthogonally-Projected EfficientNet (3D-OPEN) to effectively capture the discriminant information embedded in the 3D TRD cubes for accurate classification. The proposed model aggregates the discriminant information from three orthogonal planes projected from the 3D feature space. It alleviates the difficulty of 3D CNNs in exploiting sparse semantic abstractions directly from the high-dimensional 3D representation. The proposed method is evaluated on the Millimeter-Wave Radar Walking Dataset. It significantly and consistently outperforms the state-of-the-art methods for radar activity recognition.