Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Stochastic Adversarial Video Prediction
| Date | Stars |
|---|---|
| 2026-07-24 | 305 |
| 2026-07-25 | 305 |
| 2026-07-28 | 305 |
| 2026-07-30 | 305 |
| 2026-08-06 | 305 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# Stochastic Adversarial Video Prediction [[Project Page]](https://alexlee-gk.github.io/video_prediction/) [[Paper]](https://arxiv.org/abs/1804.01523) TensorFlow implementation for stochastic adversarial video prediction. Given a sequence of initial frames, our model is able to predict future frames of various possible futures. For example, in the next two sequences, we show the ground truth sequence on the left and random predictions of our model on the right. Predicted frames are indicated by the yellow bar at the bottom. For more examples, visit the [project page](https://alexlee-gk.github.io/video_prediction/). <img src="https://alexlee-gk.github.io/video_prediction/index_files/images/bair_action_free_random_00066_crop.gif" height="96"> <img src="https://alexlee-gk.github.io/video_prediction/index_files/images/bair_action_free_random_00006_crop.gif" height="96"> **Stochastic Adversarial Video Prediction,** [Alex X. Lee](https://people.eecs.berkeley.edu/~alexlee_gk/), [Richard Zhang](https://richzhang.github.io/), [Frederik Ebert](https://febert.github.io/), [Pieter Abbeel](https://people.eecs.berkeley.edu/~pabbeel/), [Chelsea Finn](https://people.eecs.berkeley.edu/~cbfinn/), [Sergey Levine](https://people.eecs.berkeley.edu/~svlevine/). arXiv preprint arXiv:1804.01523, 2018. An alternative implementation of SAVP is available in the [Tensor2Tensor](https://github.com/tensorflow/tensor2tensor) library. ## Getting Started ### ### Prerequisites - Linux or macOS - Python 3 - CPU or NVIDIA GPU + CUDA CuDNN ### Installation - Clone this repo: ```bash git clone -b master --single-branch https://github.com/alexlee-gk/video_prediction.git cd video_prediction ``` - Install TensorFlow >= 1.9 and dependencies from http://tensorflow.org/ - Install ffmpeg (optional, used to generate GIFs for visualization, e.g. in TensorBoard) - Install other dependencies ```bash pip install -r requirements.txt ``` ### Miscellaneous installation considerations - In python >= 3.6, make sure to add the root directory to the `PYTHONPATH`, e.g. `export PYTHONPATH=path/to/video_prediction`. - For the best speed and experimental results, we recommend using cudnn version 7.3.0.29 and any tensorflow version >= 1.9 and <= 1.12. The final training loss is worse when using cudnn versions 7.3.1.20 or 7.4.1.5, compared to when using versions 7.3.0.29 and below. - In macOS, make sure that bash >= 4.0 is used (needed for associative arrays in `download_model.sh` script). ### Use a Pre-trained Model - Download and preprocess a dataset (e.g. `bair`): ```bash bash data/download_and_preprocess_dataset.sh bair ``` - Download a pre-trained model (e.g. `ours_savp`) for the action-free version of that dataset (i.e. `bair_action_free`): ```bash bash pretrained_models/download_model.sh bair_action_free ours_savp ``` - Sample predictions from the model: ```bash CUDA_VISIBLE_DEVICES=0 python scripts/generate.py --input_dir data/bair \ --dataset_hparams sequence_length=30 \ --checkpoint pretrained_models/bair_action_free/ours_savp \ --mode test \ --results_dir results_test_samples/bair_action_free ``` - The predictions are saved as images and GIFs in `results_test_samples/bair_action_free/ours_savp`. - Evaluate predictions from the model using full-reference metrics: ```bash CUDA_VISIBLE_DEVICES=0 python scripts/evaluate.py --input_dir data/bair \ --dataset_hparams sequence_length=30 \ --checkpoint pretrained_models/bair_action_free/ours_savp \ --mode test \ --results_dir results_test/bair_action_free ``` - The results are saved in `results_test/bair_action_free/ours_savp`. - See evaluation details of our experiments in [`scripts/generate_all.sh`](scripts/generate_all.sh) and [`scripts/evaluate_all.sh`](scripts/evaluate_all.sh). ### Model Training - To train a model, download and preprocess a dataset (e.g. `bair`): ```bash bash data/download_and_preprocess_dataset.sh bair ``` - Train a model (e.g. our SAVP model on the BAIR action-free robot pushing datas
Excerpt of 8,351 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:685e9f77b412633d, topic:gan
matched fp:685e9f77b412633d, topic:video-generation