Convolutional Point Transformer

Kaul, C., Mitton, J., Dai, H. and Murray-Smith, R. (2023) Convolutional Point Transformer. In: 16th Asian Conference on Computer Vision (ACCV2022), Macau SAR, China, 4-8 December 2022, pp. 308-324. ISBN 9783031270659 (doi: 10.1007/978-3-031-27066-6_22)

[img] Text
282737.pdf - Accepted Version
Restricted to Repository staff only until 9 March 2024.



We present CpT: Convolutional point Transformer – a novel neural network layer for dealing with the unstructured nature of 3D point cloud data. CpT is an improvement over existing MLP and convolution layers for point cloud processing, as well as existing 3D point cloud processing transformer layers. It achieves this feat due to its effectiveness in creating a novel and robust attention-based point set embedding through a convolutional projection layer crafted for processing dynamically local point set neighbourhoods. The resultant point set embedding is robust to the permutations of the input points. Our novel layer builds over local neighbourhoods of points obtained via a dynamic graph computation at each layer of the network’s structure. It is fully differentiable and can be stacked just like convolutional layers to learn intrinsic properties of the points. Further, we propose a novel Adaptive Global Feature layer that learns to aggregate features from different representations into a better global representation of the point cloud. We evaluate our models on standard benchmark ModelNet40 classification and ShapeNet part segmentation datasets to show that our layer can serve as an effective addition for various point cloud processing tasks while effortlessly integrating into existing point cloud processing architectures to provide significant performance boosts.

Item Type:Conference Proceedings
Additional Information:Chaitanya Kaul and Roderick Murray-Smith acknowledge funding from the QuantIC project funded by the EPSRC Quantum Technology Programme (grant EP/M01326X/1) and the iCAIRD project, funded by Innovate UK (project number 104690). Joshua Mitton is supported by a University of Glasgow Lord Kelvin Adam Smith Studentship. Roderick Murray-Smith acknowledges funding support from EPSRC grant EP/R018634/1, Closed-loop Data Science.
Glasgow Author(s) Enlighten ID:Murray-Smith, Professor Roderick and Mitton, Mr Joshua and Kaul, Dr Chaitanya
Authors: Kaul, C., Mitton, J., Dai, H., and Murray-Smith, R.
College/School:College of Science and Engineering > School of Chemistry
College of Science and Engineering > School of Computing Science
College of Science and Engineering > School of Engineering
Copyright Holders:Copyright © 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
First Published:First published in ACCV 2022, LNCS 13848
Publisher Policy:Reproduced in accordance with the publisher copyright policy
Related URLs:

University Staff: Request a correction | Enlighten Editors: Update this record

Project CodeAward NoProject NamePrincipal InvestigatorFunder's NameFunder RefLead Dept
190841UK Quantum Technology Hub in Enhanced Quantum ImagingMiles PadgettEngineering and Physical Sciences Research Council (EPSRC)EP/M01326X/1P&S - Physics & Astronomy
304546I-CAIRD: Industrial Centre for AI Research in Digital DiagnosticsKeith MuirInnovate UK (INNOVATE)104690Stroke & Brain Imaging
300982Exploiting Closed-Loop Aspects in Computationally and Data Intensive AnalyticsRoderick Murray-SmithEngineering and Physical Sciences Research Council (EPSRC)EP/R018634/1Computing Science