Implementation of a self-supervised learning model trained with contrastive and reconstruction losses on the CIFAR-10 dataset.
This repository contains the code for a self-supervised learning model trained with contrastive and reconstruction losses. The model is implemented using the CIFAR-10 dataset and an encoder-decoder architecture.
- Objective: Implement a self-supervised learning approach using contrastive and reconstruction losses.
- Data: CIFAR-10 dataset
- Model Architecture: Encoder-decoder with contrastive and reconstruction losses
- Training Procedure: Simultaneous training with pairs of images
- Evaluation: Analysis of learned representations and effectiveness of contrastive and reconstruction losses
- Future Steps: Further experiments, application of learned representations for downstream tasks
selfsupervised_contrastive_reconstruction.ipynb
: Jupyter Notebook containing the project codeREADME.md
: This file providing an overview of the project
- Clone the repository:
git clone https://github.com/chibbss/self-supervised-learning-contrastive-reconstruction.git