Tensor to Tensor 🎯

The task of TensortoTensor is a generic task where the input and output are both tensors.

📌 Inputs:

• SampleTensor [input]:
  • A tensor ranging from 1D to 5D.

• SampleTensor [target]: (optional)
  • A tensor ranging from 1D to 5D.

• Additional parameters: (optional)
  • Any other parameters or configurations specific to the transformation model or method being applied.

🛠️ Use Cases:

• Sensor Harmonization: Aligning satellite data from different sensors to ensure consistency and comparability across various satellite missions.
• Atmospheric Correction: Adjusting for atmospheric distortions in satellite imagery to enhance the clarity and accuracy of the data.
• Feature Enhancement: Highlighting specific geographical or infrastructural features in satellite data to facilitate easier identification and analysis.

🔍 Example:

import mlstac

name = "https://huggingface.co/datasets/mlstac/tensor_to_tensor_demo.json"
dataset = mlstac.dataset(name, streaming=True, framework="torch")
print(next(iter(dataset)))

# Output example:
# {'source': tensor([[[[...]]]]), 'target': tensor([[[[...]]]])}