'Is there a way to track multiple moving objects(spermatozoa) in a microscopic video without annotated dataset to train the object detetion model?

I am doing a research on "Motility Analysis of Human Spermatozoa". I have gathered 100 spermatozoa video samples of 100 persons. But this dataset is not annotated(labeled). Is there a way to build a unsupervised deep learning model with dataset?

Sample Image Frame of a video

deep-learningobject-detectionunsupervised-learningobject-tracking


Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Source: Stack Overflow

Solution Source

Related Questions

How to fix "ValueError: Expected input batch_size (1) to match target batch_size (4)."?

Validation loss and loss stuck at 0

Faster way to do multiple embeddings in PyTorch?

How to test my trained Tensor Flow model

Convert a SMILES dataset to graph

Tensorflow TypeError: Fetch argument None has invalid type <type 'N

RuntimeError: Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor) should be the same

Whenever i try to print classification report, in accuracy column it prints some other value 0.50 but my accuracy is 0.96

Are modern CNN (convolutional neural network) as DetectNet rotate invariant?

numpy random choice in Tensorflow

Keras: network doesn't train with fit_generator()

How to Merge two CNN models?

Is it possible to automatically infer the class_weight from flow_from_directory in Keras?

How to properly initialize TensorFlow GRU-layer with noisy states?

making GRU/LSTM states trainable in Tensorflow/Keras and add some random noise