PC Parts Images Dataset [Classification]

About Dataset

The images are in the ImageNet structure, with each class having its own folder containing the respective images. The images have a resolution of 256×256 pixels.

Dataset Details:

• Total number of classes: 14
• Total number of images: 3279
• Resolution: 256×256 pixels
• Image format: JPG

Data Collection Methodology:

To make this dataset:

• I looked for pictures of each PC part on Google Images and got the links.
• Then, I downloaded these images from their original sources and changed them to JPG format, making sure they were 256 pixels in size.
• While doing this, most images became smaller, but only a few got bigger.
• Lastly, I carefully checked all the images and removed any that weren’t suitable for classifying images.

Potential Task Ideas:

1. Teach a computer to recognize different PC parts in images using well-known methods like ViT, ResNet, or EfficientNet.
2. Make use of models that have already learned things from other tasks and apply that knowledge to this dataset.
3. Try out different ways to change the images a bit to see if it helps the computer learn better.
4. Make small adjustments to existing models to make them better at figuring out what’s in the images.
5. See how different methods compare in terms of how well they work on this dataset.
6. Use this dataset to see how good new ways of recognizing images are.

This dataset is sourced from Kaggle.