Advancing Trans-Domain Classification With Knowledge Distillation: Bridging LIDAR and Image Data
Recent advancements in deep learning have significantly improved image classification models, yet extending these models to alternative data forms, such as point clouds from Light Detection and Ranging (LiDAR) sensors, presents considerable challenges. This paper explores applying knowledge distilla...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10843695/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832575572816953344 |
|---|---|
| author | Jesus Eduardo Ortiz Werner Creixell |
| author_facet | Jesus Eduardo Ortiz Werner Creixell |
| author_sort | Jesus Eduardo Ortiz |
| collection | DOAJ |
| description | Recent advancements in deep learning have significantly improved image classification models, yet extending these models to alternative data forms, such as point clouds from Light Detection and Ranging (LiDAR) sensors, presents considerable challenges. This paper explores applying knowledge distillation techniques as a solution, aiming to transfer the learned competencies from established image classification frameworks to point cloud classification tasks. Our methodology involves distilling complex model insights into more computationally efficient forms suitable for LIDAR data, thus enabling substantial resource savings without sacrificing performance. Experimental evaluations across various benchmark datasets demonstrate that our distilled models not only rival their original counterparts in accuracy but also surpass conventional point cloud classification methods in both efficiency and scalability. Additionally, we delve into the impact of varying distillation techniques on model adaptability and performance within the LIDAR domain. The findings underscore the utility of knowledge distillation in enhancing the trans-domain applicability of image classification models, potentially revolutionizing their deployment across diverse data types. |
| format | Article |
| id | doaj-art-2f8d18c632a34afb823f7d8a0d4e4248 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-2f8d18c632a34afb823f7d8a0d4e42482025-01-31T23:05:18ZengIEEEIEEE Access2169-35362025-01-0113205742058310.1109/ACCESS.2025.353044510843695Advancing Trans-Domain Classification With Knowledge Distillation: Bridging LIDAR and Image DataJesus Eduardo Ortiz0https://orcid.org/0009-0009-0358-3284Werner Creixell1https://orcid.org/0000-0002-6647-6429Departamento de Electrónica, Universidad Técnica Federico Santa María, Valparaíso, ChileDepartamento de Electrónica, Universidad Técnica Federico Santa María, Valparaíso, ChileRecent advancements in deep learning have significantly improved image classification models, yet extending these models to alternative data forms, such as point clouds from Light Detection and Ranging (LiDAR) sensors, presents considerable challenges. This paper explores applying knowledge distillation techniques as a solution, aiming to transfer the learned competencies from established image classification frameworks to point cloud classification tasks. Our methodology involves distilling complex model insights into more computationally efficient forms suitable for LIDAR data, thus enabling substantial resource savings without sacrificing performance. Experimental evaluations across various benchmark datasets demonstrate that our distilled models not only rival their original counterparts in accuracy but also surpass conventional point cloud classification methods in both efficiency and scalability. Additionally, we delve into the impact of varying distillation techniques on model adaptability and performance within the LIDAR domain. The findings underscore the utility of knowledge distillation in enhancing the trans-domain applicability of image classification models, potentially revolutionizing their deployment across diverse data types.https://ieeexplore.ieee.org/document/10843695/Knowledge distillationtrans-domain classificationpoint cloud classificationLIDAR data processingdeep learningmodel compression |
| spellingShingle | Jesus Eduardo Ortiz Werner Creixell Advancing Trans-Domain Classification With Knowledge Distillation: Bridging LIDAR and Image Data IEEE Access Knowledge distillation trans-domain classification point cloud classification LIDAR data processing deep learning model compression |
| title | Advancing Trans-Domain Classification With Knowledge Distillation: Bridging LIDAR and Image Data |
| title_full | Advancing Trans-Domain Classification With Knowledge Distillation: Bridging LIDAR and Image Data |
| title_fullStr | Advancing Trans-Domain Classification With Knowledge Distillation: Bridging LIDAR and Image Data |
| title_full_unstemmed | Advancing Trans-Domain Classification With Knowledge Distillation: Bridging LIDAR and Image Data |
| title_short | Advancing Trans-Domain Classification With Knowledge Distillation: Bridging LIDAR and Image Data |
| title_sort | advancing trans domain classification with knowledge distillation bridging lidar and image data |
| topic | Knowledge distillation trans-domain classification point cloud classification LIDAR data processing deep learning model compression |
| url | https://ieeexplore.ieee.org/document/10843695/ |
| work_keys_str_mv | AT jesuseduardoortiz advancingtransdomainclassificationwithknowledgedistillationbridginglidarandimagedata AT wernercreixell advancingtransdomainclassificationwithknowledgedistillationbridginglidarandimagedata |