{"id":24437216,"url":"https://github.com/chaganti-reddy/variational-autoencoder","last_synced_at":"2025-03-13T23:09:00.437Z","repository":{"id":231464336,"uuid":"781808414","full_name":"Chaganti-Reddy/Variational-AutoEncoder","owner":"Chaganti-Reddy","description":"Simple implementation of Variational Autoencoder on MNIST with keras number Dataset ","archived":false,"fork":false,"pushed_at":"2024-09-09T03:38:39.000Z","size":5686,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2024-09-09T04:54:50.584Z","etag":null,"topics":["cnn-keras","encoder-decoder","noise-reduction","variational-autoencoder"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Chaganti-Reddy.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2024-04-04T04:38:32.000Z","updated_at":"2024-09-09T03:38:42.000Z","dependencies_parsed_at":"2024-10-20T11:57:40.449Z","dependency_job_id":null,"html_url":"https://github.com/Chaganti-Reddy/Variational-AutoEncoder","commit_stats":null,"previous_names":["chaganti-reddy/variational-autoencoder"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Chaganti-Reddy%2FVariational-AutoEncoder","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Chaganti-Reddy%2FVariational-AutoEncoder/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Chaganti-Reddy%2FVariational-AutoEncoder/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Chaganti-Reddy%2FVariational-AutoEncoder/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Chaganti-Reddy","download_url":"https://codeload.github.com/Chaganti-Reddy/Variational-AutoEncoder/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":243495502,"owners_count":20299922,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["cnn-keras","encoder-decoder","noise-reduction","variational-autoencoder"],"created_at":"2025-01-20T18:15:46.563Z","updated_at":"2025-03-13T23:09:00.407Z","avatar_url":"https://github.com/Chaganti-Reddy.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Variational Autoencoder \n\n![](Images/main_gif.gif) \n   \n### Autoencoder             \n      \n- neural network with unsupervised machine-learning algorithm apply back-prop to set target value to the input  \n- auto-encoder prefers over PCA  because it can learn non-linear transformations with non-linear activation functions. more efficient to learn several layer with auto-encoder then one huge transformation with PCA.\n\n### Autoencoder Applications\n\n- Image coloring (Black-white images -\u003e colored)\n- Feature variation (Extract required feature)\n- Dimensionality Reduction\n- Denosing image (Remove Noise)\n- Remove watermark\n\n### Autoencoder Architecture\n\n- Encoder : part of NN compress the input into latent space representation\n- code : part of NN represents compressed input\n- Decoder : Decode the encoded data to original dimension\n\n### Properties of Autoencoder\n\n- Data-specific: Autoencoders are only able to meaningfully compress data similar to what they have been trained on.\n- Lossy: de-compressed output will be degrad compared to the original input\n- Unsupervised: Autoencoders are considered an unsupervised learning technique since they don’t need explicit labels to train on. But to be more precise they are self-supervised because they generate their own labels from the training data.\n\n### Types of Autoencoder\n\n1. Denoising autoencoder.\n2. Sparse Autoencoder.\n3. Deep Autoencoder.\n4. Contractive Autoencoder.\n5. Undercomplete Autoencoder.\n6. Convolutional Autoencoder.\n7. Variational Autoencoder.\n\n# convolutional Variational Autoencoder (Mnist)\n\n### Model architecture\n\n\u003cimg src=\"Images/L2_E30/variational_encoder_L2_E_30.png\" height=\"1000\"\u003e\n\n### Generated Numbers from latent space\n\n\u003cimg src=\"Images/generated_number.gif\" width=\"500\"\u003e\n\n### Encoded Dimension of Latent space from 2 to 10\n\n\u003cimg src=\"Images/Encoded_dimension.gif\" width=\"500\"\u003e\n\n### Remove Noise and Prediction\n\n|                        |                      **Latent space 2**                       |\n| :--------------------: | :-----------------------------------------------------------: |\n|  **Original Numbers**  |         \u003cimg src=\"Images/L2_E30/original_images.png\"\u003e         |\n| **Numbers prediction** | \u003cimg src=\"Images/L2_E30/prediction_from_original_images.png\"\u003e |\n|   **Noisy Numbers**    |          \u003cimg src=\"Images/L2_E30/Noisy_images.png\"\u003e           |\n|    **Remove Noise**    |   \u003cimg src=\"Images/L2_E30/prediction_from_noisy_image.png\"\u003e   |\n\n|                        |                      **Latent space 3**                       |\n| :--------------------: | :-----------------------------------------------------------: |\n|  **Original Numbers**  |         \u003cimg src=\"Images/L3_E30/original_images.png\"\u003e         |\n| **Numbers prediction** | \u003cimg src=\"Images/L3_E30/prediction_from_original_images.png\"\u003e |\n|   **Noisy Numbers**    |          \u003cimg src=\"Images/L3_E30/Noisy_images.png\"\u003e           |\n|    **Remove Noise**    |   \u003cimg src=\"Images/L3_E30/prediction_from_noisy_image.png\"\u003e   |\n\n|                        |                      **Latent space 4**                       |\n| :--------------------: | :-----------------------------------------------------------: |\n|  **Original Numbers**  |         \u003cimg src=\"Images/L4_E30/original_images.png\"\u003e         |\n| **Numbers prediction** | \u003cimg src=\"Images/L4_E30/prediction_from_original_images.png\"\u003e |\n|   **Noisy Numbers**    |          \u003cimg src=\"Images/L4_E30/Noisy_images.png\"\u003e           |\n|    **Remove Noise**    |   \u003cimg src=\"Images/L4_E30/prediction_from_noisy_image.png\"\u003e   |\n\n|                        |                      **Latent space 5**                       |\n| :--------------------: | :-----------------------------------------------------------: |\n|  **Original Numbers**  |         \u003cimg src=\"Images/L5_E30/original_images.png\"\u003e         |\n| **Numbers prediction** | \u003cimg src=\"Images/L5_E30/prediction_from_original_images.png\"\u003e |\n|   **Noisy Numbers**    |          \u003cimg src=\"Images/L5_E30/Noisy_images.png\"\u003e           |\n|    **Remove Noise**    |   \u003cimg src=\"Images/L5_E30/prediction_from_noisy_image.png\"\u003e   |\n\n|                        |                      **Latent space 6**                       |\n| :--------------------: | :-----------------------------------------------------------: |\n|  **Original Numbers**  |         \u003cimg src=\"Images/L6_E30/original_images.png\"\u003e         |\n| **Numbers prediction** | \u003cimg src=\"Images/L6_E30/prediction_from_original_images.png\"\u003e |\n|   **Noisy Numbers**    |          \u003cimg src=\"Images/L6_E30/Noisy_images.png\"\u003e           |\n|    **Remove Noise**    |   \u003cimg src=\"Images/L6_E30/prediction_from_noisy_image.png\"\u003e   |\n\n|                        |                      **Latent space 7**                       |\n| :--------------------: | :-----------------------------------------------------------: |\n|  **Original Numbers**  |         \u003cimg src=\"Images/L7_E30/original_images.png\"\u003e         |\n| **Numbers prediction** | \u003cimg src=\"Images/L7_E30/prediction_from_original_images.png\"\u003e |\n|   **Noisy Numbers**    |          \u003cimg src=\"Images/L7_E30/Noisy_images.png\"\u003e           |\n|    **Remove Noise**    |   \u003cimg src=\"Images/L7_E30/prediction_from_noisy_image.png\"\u003e   |\n\n|                        |                      **Latent space 8**                       |\n| :--------------------: | :-----------------------------------------------------------: |\n|  **Original Numbers**  |         \u003cimg src=\"Images/L8_E30/original_images.png\"\u003e         |\n| **Numbers prediction** | \u003cimg src=\"Images/L8_E30/prediction_from_original_images.png\"\u003e |\n|   **Noisy Numbers**    |          \u003cimg src=\"Images/L8_E30/Noisy_images.png\"\u003e           |\n|    **Remove Noise**    |   \u003cimg src=\"Images/L8_E30/prediction_from_noisy_image.png\"\u003e   |\n\n|                        |                      **Latent space 9**                       |\n| :--------------------: | :-----------------------------------------------------------: |\n|  **Original Numbers**  |         \u003cimg src=\"Images/L9_E30/original_images.png\"\u003e         |\n| **Numbers prediction** | \u003cimg src=\"Images/L9_E30/prediction_from_original_images.png\"\u003e |\n|   **Noisy Numbers**    |          \u003cimg src=\"Images/L9_E30/Noisy_images.png\"\u003e           |\n|    **Remove Noise**    |   \u003cimg src=\"Images/L9_E30/prediction_from_noisy_image.png\"\u003e   |\n\n|                        |                      **Latent space 10**                       |\n| :--------------------: | :------------------------------------------------------------: |\n|  **Original Numbers**  |         \u003cimg src=\"Images/L10_E30/original_images.png\"\u003e         |\n| **Numbers prediction** | \u003cimg src=\"Images/L10_E30/prediction_from_original_images.png\"\u003e |\n|   **Noisy Numbers**    |          \u003cimg src=\"Images/L10_E30/Noisy_images.png\"\u003e           |\n|    **Remove Noise**    |   \u003cimg src=\"Images/L10_E30/prediction_from_noisy_image.png\"\u003e   |\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fchaganti-reddy%2Fvariational-autoencoder","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fchaganti-reddy%2Fvariational-autoencoder","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fchaganti-reddy%2Fvariational-autoencoder/lists"}