OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Dancenet

Dancenet is an open-source Python and Keras project that generates dance-sequence video with a variational autoencoder, an LSTM and a mixture-density layer. Its repository includes a notebook and scripts for encoding images, training the model and generating or testing video.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Dancenet — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Dancenet when you want to explore a documented dance-image-to-video workflow or study how a variational autoencoder and LSTM with a mixture-density layer are used in sequence generation.

Who it is for

  • Developers exploring sequence-based image generation with Python and Keras.
  • People who want to experiment with dance-video generation using pretrained weights.
  • Researchers or learners investigating variational autoencoders and LSTM-based sequence models.

Documented uses

  • Run the notebook with the linked trained weights.
  • Encode ordered dance images into latent representations.
  • Train the model using labeled dance-sequence images.
  • Generate or test video from dance sequences.
Before you choose

Requirements & limits

The code repository is public and identifies an MIT license. The README also describes running the notebook in a FloydHub workspace with a trained-weights dataset attached.

  • The README specifies Python 3.5.2 and fixed versions of Keras, scikit-learn, NumPy and OpenCV.
  • The training workflow expects ordered image files in imgs/, named sequentially.
  • The project provides distinct steps for encoding images, testing decoded video and notebook-based training and generation.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Run the notebook with pretrained weights

Set up the documented local notebook workflow using the project's pretrained weights.

Before you start

  • Python 3.5.2
  • Keras 2.2.0, scikit-learn 0.19.1, NumPy 1.14.3 and OpenCV 3.4.1
  • The Dancenet repository and its linked trained weights

Steps

  1. Download the trained weights linked in the repository README.

  2. Extract the weights into the dancenet project directory.

  3. Run the dancegen.ipynb notebook.

Expected outcome

The documented notebook workflow for using Dancenet with pretrained weights.

Guide limitations

  • The README does not provide further notebook execution details.
  • The listed dependency versions are specific and dated.
Documentation 1 ↗
Source-based practical guide

Prepare image sequences for training

Prepare ordered dance images and follow the repository's documented from-scratch training stages.

Before you start

  • Python 3.5.2
  • Keras 2.2.0, scikit-learn 0.19.1, NumPy 1.14.3 and OpenCV 3.4.1
  • Dance-sequence images saved in an imgs/ folder and named in order, such as 1.jpg, 2.jpg

Steps

  1. Place the ordered image files in the imgs/ folder.

  2. Run model.py.

  3. Run gen_lv.py to encode the images.

  4. Run video_from_lv.py to test decoded video.

  5. Use dancegen.ipynb for the documented training and video-generation stage.

Expected outcome

A documented sequence of stages for training Dancenet from labeled image sequences and generating video.

Guide limitations

  • The README does not specify image dimensions, sequence length or further dataset preparation.
  • It does not document the expected results or detailed operation of the scripts.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
17 October 2026
Original website title
jsn5/dancenet · Official homepage title · retained in its source language
How we review sourcesSuggest a correction