CPWC:Contextual Point Wise Convolution for Object Recognition
Saved in:
| Title: | CPWC:Contextual Point Wise Convolution for Object Recognition |
|---|---|
| Authors: | Mazumder, Pratik, Singh, Pravendra, Namboodiri, Vinay |
| Source: | Mazumder, P, Singh, P & Namboodiri, V 2020, CPWC : Contextual Point Wise Convolution for Object Recognition. in 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings., 9054205, ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, vol. 2020-May, IEEE, pp. 4152-4156, 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020, Barcelona, Spain, 4/05/20. https://doi.org/10.1109/ICASSP40776.2020.9054205 |
| Publisher Information: | IEEE |
| Publication Year: | 2020 |
| Subject Terms: | convolutional neural network, deep learning, object recognition, Pointwise convolution, /dk/atira/pure/subjectarea/asjc/1700/1712, name=Software, /dk/atira/pure/subjectarea/asjc/1700/1711, name=Signal Processing, /dk/atira/pure/subjectarea/asjc/2200/2208, name=Electrical and Electronic Engineering |
| Description: | Convolutional layers are a major driving force behind the successes of deep learning. Pointwise convolution (PWC) is a 1 × 1 convolutional filter that is primarily used for parameter reduction. However, the PWC ignores the spatial information around the points it is processing. This design is by choice, in order to reduce the overall parameters and computations. However, we hypothesize that this shortcoming of PWC has a significant impact on the network performance. We propose an alternative design for pointwise convolution, which uses spatial information from the input efficiently. Our design significantly improves the performance of the networks without substantially increasing the number of parameters and computations. We experimentally show that our design results in significant improvement in the performance of the network for classification as well as detection. |
| Document Type: | article in journal/newspaper |
| File Description: | application/pdf |
| Language: | English |
| ISBN: | 978-1-5090-6631-5 1-5090-6631-4 |
| Relation: | info:eu-repo/semantics/altIdentifier/isbn/9781509066315; urn:ISBN:9781509066315 |
| DOI: | 10.1109/ICASSP40776.2020.9054205 |
| Availability: | https://researchportal.bath.ac.uk/en/publications/59ab8ace-eae8-4038-8982-dcfd6d5e3be4 https://doi.org/10.1109/ICASSP40776.2020.9054205 https://purehost.bath.ac.uk/ws/files/215043362/ICASSP20_CPWC_Contextual_Point_Wise_Convolution_for_Object_Recognition.pdf https://www.scopus.com/pages/publications/85089237978 |
| Rights: | info:eu-repo/semantics/openAccess |
| Accession Number: | edsbas.FD1CCCA |
| Database: | BASE |
Full Text Finder 
Nájsť tento článok vo Web of Science | Abstract: | Convolutional layers are a major driving force behind the successes of deep learning. Pointwise convolution (PWC) is a 1 × 1 convolutional filter that is primarily used for parameter reduction. However, the PWC ignores the spatial information around the points it is processing. This design is by choice, in order to reduce the overall parameters and computations. However, we hypothesize that this shortcoming of PWC has a significant impact on the network performance. We propose an alternative design for pointwise convolution, which uses spatial information from the input efficiently. Our design significantly improves the performance of the networks without substantially increasing the number of parameters and computations. We experimentally show that our design results in significant improvement in the performance of the network for classification as well as detection. |
|---|---|
| ISBN: | 9781509066315 1509066314 |
| DOI: | 10.1109/ICASSP40776.2020.9054205 |