Module: tfm.vision.configs.video_classification

Video classification configuration definition.

Classes

class DataConfig: The base configuration for building datasets.

class Losses

class Metrics

class VideoClassificationModel: The model config.

class VideoClassificationTask: The task config.

Functions

add_trainer(...): Add and config a trainer to the experiment config.

kinetics400(...): Generated Kinetics 400 dataset configs.

kinetics600(...): Generated Kinetics 600 dataset configs.

kinetics700(...): Generated Kinetics 600 dataset configs.

kinetics700_2020(...): Generated Kinetics 600 dataset configs.

video_classification(...): Video classification general.

video_classification_kinetics400(...): Video classification on Kinetics 400 with resnet.

video_classification_kinetics600(...): Video classification on Kinetics 600 with resnet.

video_classification_kinetics700(...): Video classification on Kinetics 700 with resnet.

video_classification_kinetics700_2020(...): Video classification on Kinetics 700 2020 with resnet.

video_classification_ucf101(...): Video classification on UCF-101 with resnet.