Effect of Camera Choice on Image-Classification Inference
The field of image classification using Convolutional Neural Networks (CNNs) to predict the principal object in an image has seen many recent innovations. One aspect that has not been extensively explored is the effect of the camera employed to acquire images for inference. We investigate this by ca...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/15/1/246 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841549399911563264 |
|---|---|
| author | Jason Brown Andy Nguyen Nawin Raj |
| author_facet | Jason Brown Andy Nguyen Nawin Raj |
| author_sort | Jason Brown |
| collection | DOAJ |
| description | The field of image classification using Convolutional Neural Networks (CNNs) to predict the principal object in an image has seen many recent innovations. One aspect that has not been extensively explored is the effect of the camera employed to acquire images for inference. We investigate this by capturing comparable images of five drinking vessels using six cameras in various scenarios. We examine the classification ranking of object classes when these images are input to an independently pretrained Resnet-18 model based on the ImageNet-1k dataset. We find that the camera used can affect the top prediction of object class, particularly in scenarios with a more complex background. This is the case even when the cameras have similar fields of view. We also introduce a metric called selectivity, defined as the mean absolute difference between prediction probabilities of similar relevant object classes (such as cups and mugs). We show that the effect of the camera is largest when the selectivity of the pretrained model between these object classes is small. The effect of camera choice is also demonstrated quantitatively by examining Cohen’s Kappa (κ) statistic. Finally, we make recommendations on mitigating the effect of the camera on image-classification inference. |
| format | Article |
| id | doaj-art-2f5cada103d446afb3833334fdaa37f3 |
| institution | Kabale University |
| issn | 2076-3417 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj-art-2f5cada103d446afb3833334fdaa37f32025-01-10T13:14:58ZengMDPI AGApplied Sciences2076-34172024-12-0115124610.3390/app15010246Effect of Camera Choice on Image-Classification InferenceJason Brown0Andy Nguyen1Nawin Raj2School of Engineering, University of Southern Queensland, Brisbane 4300, AustraliaSchool of Engineering, University of Southern Queensland, Brisbane 4300, AustraliaSchool of Mathematics, Physics and Computing, University of Southern Queensland, Brisbane 4300, AustraliaThe field of image classification using Convolutional Neural Networks (CNNs) to predict the principal object in an image has seen many recent innovations. One aspect that has not been extensively explored is the effect of the camera employed to acquire images for inference. We investigate this by capturing comparable images of five drinking vessels using six cameras in various scenarios. We examine the classification ranking of object classes when these images are input to an independently pretrained Resnet-18 model based on the ImageNet-1k dataset. We find that the camera used can affect the top prediction of object class, particularly in scenarios with a more complex background. This is the case even when the cameras have similar fields of view. We also introduce a metric called selectivity, defined as the mean absolute difference between prediction probabilities of similar relevant object classes (such as cups and mugs). We show that the effect of the camera is largest when the selectivity of the pretrained model between these object classes is small. The effect of camera choice is also demonstrated quantitatively by examining Cohen’s Kappa (κ) statistic. Finally, we make recommendations on mitigating the effect of the camera on image-classification inference.https://www.mdpi.com/2076-3417/15/1/246image classificationcomputer visioninferencepredictioncamera |
| spellingShingle | Jason Brown Andy Nguyen Nawin Raj Effect of Camera Choice on Image-Classification Inference Applied Sciences image classification computer vision inference prediction camera |
| title | Effect of Camera Choice on Image-Classification Inference |
| title_full | Effect of Camera Choice on Image-Classification Inference |
| title_fullStr | Effect of Camera Choice on Image-Classification Inference |
| title_full_unstemmed | Effect of Camera Choice on Image-Classification Inference |
| title_short | Effect of Camera Choice on Image-Classification Inference |
| title_sort | effect of camera choice on image classification inference |
| topic | image classification computer vision inference prediction camera |
| url | https://www.mdpi.com/2076-3417/15/1/246 |
| work_keys_str_mv | AT jasonbrown effectofcamerachoiceonimageclassificationinference AT andynguyen effectofcamerachoiceonimageclassificationinference AT nawinraj effectofcamerachoiceonimageclassificationinference |