https://github.com/msikorski93/product-recommendation-engine-based-on-visual-similarity
Deep learning image recognition based on visual similarity for e-commerce.
https://github.com/msikorski93/product-recommendation-engine-based-on-visual-similarity
deep-learning e-commerce image-recognition keras recommender-system resnet50 tensorflow vgg16
Last synced: 6 months ago
JSON representation
Deep learning image recognition based on visual similarity for e-commerce.
- Host: GitHub
- URL: https://github.com/msikorski93/product-recommendation-engine-based-on-visual-similarity
- Owner: msikorski93
- Created: 2023-03-20T18:43:26.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-03-20T18:53:00.000Z (over 2 years ago)
- Last Synced: 2025-04-23T11:56:48.052Z (6 months ago)
- Topics: deep-learning, e-commerce, image-recognition, keras, recommender-system, resnet50, tensorflow, vgg16
- Language: Jupyter Notebook
- Homepage:
- Size: 3.83 MB
- Stars: 7
- Watchers: 1
- Forks: 4
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Product-Recommendation-Engine-Based-on-Visual-Similarity
The subject of this repository was to develop a very basic recommender engine for fashion products. For this case, the system
should recommend goods that look similar. The idea behind is very simple: if a customer is showing interest towards a specific
product by browsing its page, he may also be interested by products that are visually alike. We will implement CNNs with an
image feature extraction approach, with Keras and TensorFlow packages. The recommender engines were
based on pre-trained models: ResNet50 and VGG16.
The recommender systems are able to find similar products accurately: most of the time the retrieved products have the same
purpose and look very similar. They certainly work well in terms of shapes; patterns are not always recognized.