# 3D Depthwise Convolution: Reducing Model Parameters in 3D Vision Tasks

arXiv: https://arxiv.org/abs/1808.01556 · HTML: https://mlsanity.com/paper/1808.01556/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Aug 2018.

## Built on by (most cited first, 9 of 9)

- [Make-A-Video: Text-to-Video Generation without Text-Video Data](https://mlsanity.com/paper/2209.14792/) (2022)
- [MobileStereoNet: Towards Lightweight Deep Networks for Stereo Matching](https://mlsanity.com/paper/2108.09770/) (2021)
- [Collaborative Semantic Occupancy Prediction with Hybrid Feature Fusion in Connected Automated Vehicles](https://mlsanity.com/paper/2402.07635/) (2024)
- [Turbo-VAED: Fast and Stable Transfer of Video-VAEs to Mobile Devices](https://mlsanity.com/paper/2508.09136/) (2025)
- [PocketNet: A Smaller Neural Network for Medical Image Analysis](https://mlsanity.com/paper/2104.10745/) (2021)
- [Separable Convolutions for Optimizing 3D Stereo Networks](https://mlsanity.com/paper/2108.10216/) (2021)
- [Med-Tuning: A New Parameter-Efficient Tuning Framework for Medical Volumetric Segmentation](https://mlsanity.com/paper/2304.10880/) (2023)
- [Neural Reconstruction of LiDAR Point Clouds under Jamming Attacks via Full-Waveform Representation and Simultaneous Laser Sensing](https://mlsanity.com/paper/2604.00371/) (2026)
- [SED:Lightweight Saliency prediction for Event-based data via Distillation](https://mlsanity.com/paper/2606.14631/) (2026)

## Abstract (arXiv)

Standard 3D convolution operations require much larger amounts of memory and computation cost than 2D convolution operations. The fact has hindered the development of deep neural nets in many 3D vision tasks. In this paper, we investigate the possibility of applying depthwise separable convolutions in 3D scenario and introduce the use of 3D depthwise convolution. A 3D depthwise convolution splits a single standard 3D convolution into two separate steps, which would drastically reduce the number of parameters in 3D convolutions with more than one order of magnitude. We experiment with 3D depthwise convolution on popular CNN architectures and also compare it with a similar structure called pseudo-3D convolution. The results demonstrate that, with 3D depthwise convolutions, 3D vision tasks like classification and reconstruction can be carried out with more light-weighted neural networks while still delivering comparable performances.