ujjax/pred-rnn
PredRNN: Recurrent Neural Networks for Predictive Learning using Spatiotemporal LSTMs
This tool helps researchers and deep learning practitioners explore advanced methods for predicting future video frames. It takes a sequence of video frames as input and generates a prediction of what the subsequent frames will look like. This is useful for those working on video prediction, anomaly detection in video, or general spatiotemporal sequence modeling tasks.
127 stars. No commits in the last 6 months.
Use this if you are a deep learning researcher or practitioner interested in implementing or experimenting with predictive learning models for video.
Not ideal if you are looking for an off-the-shelf solution for video prediction without needing to understand or modify deep learning model architectures.
Stars
127
Forks
39
Language
Python
License
MIT
Category
Last pushed
Oct 09, 2019
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/ujjax/pred-rnn"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
hangyav/textLSP
Language server for text spell and grammar check with various tools.
google-research/electra
ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators
gsurma/text_predictor
Char-level RNN LSTM text generatorđź“„.
oliverguhr/fullstop-deep-punctuation-prediction
A model that predicts the punctuation of English, Italian, French and German texts.
ivanliu1989/SwiftKey-Natural-language
SwiftKey, our corporate partner in this capstone, builds a smart keyboard that makes it easier...