PTLFlow

Getting started:

  • Installation
  • Predict optical flow with a pretrained model
  • Run validation on a model
  • Train an existing model
  • Predicting optical flow on test datasets
  • Using config files
  • Model Benchmarking
  • Changes from PTLFlow v0.3 to v0.4

Models:

  • List of available models
  • Checkpoint List

Datasets:

  • List of supported datasets

Results:

  • Accuracy - EPE
  • Accuracy - EPE and Fl-All
  • Model benchmark
  • Paper/PTLFlow metrics
  • Result plots

Customizing:

  • Adding a new model
  • Adding a new dataset

API:

  • train_dataset and val_dataset
  • scripts
  • ptlflow.models
  • ptflow.data
  • ptlflow.utils
    • correlation.py
    • dummy_datasets.py
    • flow_metrics.py
    • flow_utils.py
    • flowpy_torch.py
    • io_adapter.py
    • timer.py
    • utils.py
    • logger.py
PTLFlow
  • ptlflow.utils
  • View page source

ptlflow.utils

  • correlation.py
  • dummy_datasets.py
  • flow_metrics.py
  • flow_utils.py
  • flowpy_torch.py
  • io_adapter.py
  • timer.py
  • utils.py

Callbacks

  • logger.py
Previous Next

© Copyright 2024, Henrique Morimitsu.

Built with Sphinx using a theme provided by Read the Docs.