{"id":21892968,"url":"https://github.com/chunyuanli/alice","last_synced_at":"2025-04-15T15:04:48.992Z","repository":{"id":92717948,"uuid":"102434437","full_name":"ChunyuanLI/ALICE","owner":"ChunyuanLI","description":"NIPS 2017: ALICE: Towards Understanding Adversarial Learning for Joint Distribution Matching","archived":false,"fork":false,"pushed_at":"2018-08-16T23:52:23.000Z","size":73524,"stargazers_count":82,"open_issues_count":2,"forks_count":24,"subscribers_count":2,"default_branch":"master","last_synced_at":"2025-04-15T15:04:37.967Z","etag":null,"topics":["bayesian-inference","entropy","generative-adversarial-network","image-translation","latent-variable-models"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/ChunyuanLI.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2017-09-05T04:33:25.000Z","updated_at":"2024-04-13T06:40:34.000Z","dependencies_parsed_at":null,"dependency_job_id":"2aae0a4e-4689-4151-8132-b1a08d0d4a6d","html_url":"https://github.com/ChunyuanLI/ALICE","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ChunyuanLI%2FALICE","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ChunyuanLI%2FALICE/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ChunyuanLI%2FALICE/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ChunyuanLI%2FALICE/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ChunyuanLI","download_url":"https://codeload.github.com/ChunyuanLI/ALICE/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":249094932,"owners_count":21211837,"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":["bayesian-inference","entropy","generative-adversarial-network","image-translation","latent-variable-models"],"created_at":"2024-11-28T13:00:11.684Z","updated_at":"2025-04-15T15:04:48.973Z","avatar_url":"https://github.com/ChunyuanLI.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# ALICE\nAdversarially Learned Inference with Conditional Entropy (**ALICE**)\n\n[ALICE: Towards Understanding Adversarial Learning for Joint Distribution Matching](https://arxiv.org/abs/1709.01215)  \n [Chunyuan Li](http://chunyuan.li/),\n [Hao Liu](https://hliu96.github.io/), \n [Changyou Chen](https://www.cse.buffalo.edu/~changyou/), \n [Yunchen Pu](https://scholar.google.com/citations?user=ftW7RoAAAAAJ\u0026hl=en), \n [Liqun Chen](https://scholar.google.com/citations?user=T9T8Il0AAAAJ\u0026hl=en), \n [Ricardo Henao](https://scholar.google.com/citations?user=p_mm4-YAAAAJ),\n [Lawrence Carin](http://people.ee.duke.edu/~lcarin/)  \n Duke University. NIPS, 2017.\n\n![](/plot_generation/figures_alice/alice_log_movie.gif)\n\n*[Alice4Alice](https://github.com/ChunyuanLI/Alice4Alice): ALICE algorithms for painting the cartoon of Alice's Adventures in Wonderland*\n\n## Four variants of ALICE on toy datasets\nIn *unsupervised learning* case: \n\n- (a) Explicit cycle-consistency ([`ALICE_l2.py`](/toy_data/ALICE_l2.py)) \n- (b) Implicit cycle-consistency ([`ALICE_A.py`](/toy_data/ALICE_A.py))\n\nIn *weakly-supervised learning* case:\n\n- (c) Explicit mapping  ([`ALICE_l2_l2.py`](/toy_data/ALICE_l2_l2.py)) \n- (d) Implicit mapping  ([`ALICE_A_A.py`](/toy_data/ALICE_A_A.py)) \n\n## Reproduce figures in the paper\n\n[`plot_generation/alice_plots_paper.ipynb`](./plot_generation/alice_plots_paper.ipynb)\n\n## Real datasets\n\n### MNIST\nWe study the impact of weighting hyperparameter (\\lambda) for CE regularizer. The performance of image generation is evaluated by **inception score (ICP)**, and image reconstruction is evaluted by **mean square error (MSE)**.\n\nBest ICP=9.279 ± 0.07, and MSE=0.0803 ± 0.007, when \\lambda=1\n\nNote: we pre-trained a \"perfect\" MNIST classifier (100\\% training accuracy) to compute the [inception score for MNIST](https://github.com/ChunyuanLI/MNIST_Inception_Score).\n\nImage Generation             |  Image Reconstruction\n:-------------------------:|:-------------------------:\n![](/plot_generation/figures/mnist_icp_weighting.png)  |  ![](/plot_generation/figures/mnist_mse_weighting.png)\n\n### CIFAR\n\nBest ICP=6.015 ± 0.0284, and MSE=0.4155 ± 0.2015, when \\lambda=1e-6. Larger \\lambda leads to lower MSE.\n\nNote: The quality of generated cifar images is evaluated via the [inception score based on ImageNet](https://github.com/openai/improved-gan/tree/master/inception_score).\n\nImage Generation             |  Image Reconstruction\n:-------------------------:|:-------------------------:\n![](/plot_generation/figures/cifar_icp_weighting.png)  |  ![](/plot_generation/figures/cifar_mse_weighting.png)\n\n\n### CelebA\n### Car2Car\n### Edge2Shoes\n\n\n## Citation\nIf you use this code for your research, please cite our [paper](https://arxiv.org/abs/1709.01215):\n\n```\n@article{li2017alice,\n  title={ALICE: Towards Understanding Adversarial Learning for Joint Distribution Matching},\n  author={Li, Chunyuan and Liu, Hao and Chen, Changyou and Pu, Yunchen and Chen, Liqun and Henao, Ricardo and Carin, Lawrence},\n  journal={Neural Information Processing Systems (NIPS)},\n  year={2017}\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fchunyuanli%2Falice","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fchunyuanli%2Falice","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fchunyuanli%2Falice/lists"}