Heavy and Lightweight Deep Learning Models for Semantic Segmentation: A Survey
Semantic segmentation is an important computer vision task due to its numerous real-world applications such as autonomous driving, video surveillance, medical image analysis, robotics, augmented reality, among others, and its popularity increased with the development of deep learning approaches. We...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10840225/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832576775048134656 |
|---|---|
| author | Cristina Carunta Alina Carunta Calin-Adrian Popa |
| author_facet | Cristina Carunta Alina Carunta Calin-Adrian Popa |
| author_sort | Cristina Carunta |
| collection | DOAJ |
| description | Semantic segmentation is an important computer vision task due to its numerous real-world applications such as autonomous driving, video surveillance, medical image analysis, robotics, augmented reality, among others, and its popularity increased with the development of deep learning approaches. We provide a detailed review comprising the most significant methods for both heavy and lightweight two-dimensional (2D) semantic segmentation, starting with the introduction of convolutional neural networks until the use of Transformer architecture, the latter being a widely adopted model with state-of-the-art results in several artificial intelligence fields. The methods involved are described from the architectural design perspective, including encoder-decoder architectures, multi-resolution branches approaches, two-pathway encoder architectures, attention-based models, and pyramid-based models. Additionally, some of the most popular datasets and performance metrics are presented. Further, we investigate the limitations of these methods, compare their performance on Pascal VOC 2012, Cityscapes, and ADE20K datasets, and finally indicate future research directions. |
| format | Article |
| id | doaj-art-a407120564774f3eb83c7cdf5cba036f |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-a407120564774f3eb83c7cdf5cba036f2025-01-31T00:01:31ZengIEEEIEEE Access2169-35362025-01-0113177451776510.1109/ACCESS.2025.352981210840225Heavy and Lightweight Deep Learning Models for Semantic Segmentation: A SurveyCristina Carunta0https://orcid.org/0009-0003-7597-0781Alina Carunta1https://orcid.org/0009-0005-0036-5418Calin-Adrian Popa2https://orcid.org/0000-0003-4445-8091Department of Computers and Information Technology, Politehnica University of Timişoara, Timişoara, RomaniaDepartment of Computer Science, West University of Timişoara, Timişoara, RomaniaDepartment of Computers and Information Technology, Politehnica University of Timişoara, Timişoara, RomaniaSemantic segmentation is an important computer vision task due to its numerous real-world applications such as autonomous driving, video surveillance, medical image analysis, robotics, augmented reality, among others, and its popularity increased with the development of deep learning approaches. We provide a detailed review comprising the most significant methods for both heavy and lightweight two-dimensional (2D) semantic segmentation, starting with the introduction of convolutional neural networks until the use of Transformer architecture, the latter being a widely adopted model with state-of-the-art results in several artificial intelligence fields. The methods involved are described from the architectural design perspective, including encoder-decoder architectures, multi-resolution branches approaches, two-pathway encoder architectures, attention-based models, and pyramid-based models. Additionally, some of the most popular datasets and performance metrics are presented. Further, we investigate the limitations of these methods, compare their performance on Pascal VOC 2012, Cityscapes, and ADE20K datasets, and finally indicate future research directions.https://ieeexplore.ieee.org/document/10840225/Complex deep learning modelsconvolutional neural networksreal-time modelssemantic segmentationTransformer model |
| spellingShingle | Cristina Carunta Alina Carunta Calin-Adrian Popa Heavy and Lightweight Deep Learning Models for Semantic Segmentation: A Survey IEEE Access Complex deep learning models convolutional neural networks real-time models semantic segmentation Transformer model |
| title | Heavy and Lightweight Deep Learning Models for Semantic Segmentation: A Survey |
| title_full | Heavy and Lightweight Deep Learning Models for Semantic Segmentation: A Survey |
| title_fullStr | Heavy and Lightweight Deep Learning Models for Semantic Segmentation: A Survey |
| title_full_unstemmed | Heavy and Lightweight Deep Learning Models for Semantic Segmentation: A Survey |
| title_short | Heavy and Lightweight Deep Learning Models for Semantic Segmentation: A Survey |
| title_sort | heavy and lightweight deep learning models for semantic segmentation a survey |
| topic | Complex deep learning models convolutional neural networks real-time models semantic segmentation Transformer model |
| url | https://ieeexplore.ieee.org/document/10840225/ |
| work_keys_str_mv | AT cristinacarunta heavyandlightweightdeeplearningmodelsforsemanticsegmentationasurvey AT alinacarunta heavyandlightweightdeeplearningmodelsforsemanticsegmentationasurvey AT calinadrianpopa heavyandlightweightdeeplearningmodelsforsemanticsegmentationasurvey |