Upload external_waveplate pilot: README.md

README.md

@@ -1,3 +1,190 @@
 ---
-license: apache-2.0
+pretty_name: MMPD-Bench
+license: cc-by-nc-4.0
+tags:
+- polarimetry
+- mueller-matrix
+- biomedical-imaging
+- healthy-bone-cell
+- scientific-computing
+- parquet
+task_categories:
+- image-to-image
+configs:
+- config_name: external_waveplate
+  default: true
+  data_files:
+  - split: external_waveplate
+    path: data/external_waveplate-*.parquet
 ---
+
+# MMPD-Bench
+
+## Dataset Summary
+
+TODO: Describe MMPD-Bench, the scientific motivation, and the intended benchmark
+use. Mention that the dataset contains paired Mueller matrix observations and
+Lu-Chipman reference decomposition outputs.
+
+Current pilot upload: `external_waveplate` only.
+
+## Task Definition
+
+TODO: Define the task as modality fission from a Mueller matrix patch to six
+polarimetric target modalities.
+
+- Input: Mueller matrix tensor, shape `[16, H, W]`, channel-first.
+- Output: target modality tensor, shape `[6, H, W]`, channel-first.
+- Target channel order: `D`, `Delta`, `eta`, `theta`, `psi`, `R`.
+
+## Data Sources
+
+TODO: Describe healthy bone cell samples, waveplate samples, and
+multi-wavelength measurements. Include acquisition context and de-identification
+statement if applicable.
+
+## File Structure
+
+```text
+MMPD-Bench/
+├── README.md
+├── data/
+├── metadata/
+├── preview/
+└── raw/
+```
+
+TODO: Replace this with final shard counts and file sizes after conversion.
+
+## Tensor Schema
+
+Each row represents one paired patch/sample:
+
+```python
+{
+    "sample_id": str,
+    "source_id": str,
+    "split": str,
+    "subset": str,         # healthy_bone_cell, waveplate, spectral
+    "specimen_type": str,  # healthy_bone_cell, waveplate, spectral, unknown
+    "wavelength_nm": int,
+    "mueller_shape": list[int],
+    "target_shape": list[int],
+    "mueller": array,  # float32, channel-first
+    "target": array,   # float32, channel-first
+}
+```
+
+TODO: Confirm whether arrays are stored as physical values, normalized values,
+or both. The `external_waveplate` pilot shard currently stores tensors with
+shapes `[16, 200, 200]` and `[6, 200, 200]`.
+
+## Channel Conventions
+
+Mueller channel order:
+
+```text
+M11, M12, M13, M14,
+M21, M22, M23, M24,
+M31, M32, M33, M34,
+M41, M42, M43, M44
+```
+
+Target channel order:
+
+```text
+D, Delta, eta, theta, psi, R
+```
+
+TODO: Document the mapping from local filenames such as `Ita`, `ita`, or `Eta`
+to public channel name `eta`.
+
+## Physical Parameter Definitions
+
+TODO: Define every target parameter and confirm ranges.
+
+```text
+D, Delta: [0, 1]
+eta, R: [0, pi)
+theta, psi: [-pi/2, pi/2)
+```
+
+## Reference Label Generation
+
+The target modalities are generated using Lu-Chipman decomposition from the
+measured Mueller matrices. They should be interpreted as reference solver
+outputs for benchmarking surrogate models and physics consistency, rather than
+direct human annotations or absolute ground truth for biological tissue.
+
+TODO: Add decomposition implementation details, preprocessing, and any numerical
+stability handling.
+
+## Splits
+
+TODO: Fill exact counts after conversion.
+
+| Split | Subset | Samples | Notes |
+|---|---:|---:|---|
+| train | healthy_bone_cell | TODO | Primary training split |
+| validation | healthy_bone_cell | TODO | Model selection split |
+| test | healthy_bone_cell | TODO | Clean healthy bone cell test |
+| external_waveplate | waveplate | 18 | External generalisation at 633 nm |
+| external_spectral_610 | spectral | TODO | External wavelength test |
+| external_spectral_650 | spectral | TODO | External wavelength test |
+| external_spectral_690 | spectral | TODO | External wavelength test |
+
+TODO: State whether healthy bone cell splits are specimen-level or patch-level.
+If patch-level, describe leakage risk as a limitation.
+
+## External Test Sets
+
+TODO: Describe waveplate and spectral external test sets, including wavelength,
+sample source, and intended usage.
+
+## Benchmark Protocols
+
+1. Clean healthy bone cell test: train on `train`, tune on `validation`, report on
+   `test`.
+2. Noisy healthy bone cell robustness: evaluate the test set with
+   `sigma_noise = 0.1 * sigma_pixel`.
+3. External generalisation: train only on healthy bone cell training data and
+   report on `external_waveplate` and each `external_spectral_*` split.
+
+## Metrics
+
+TODO: List recommended metrics for each target modality and any aggregate score.
+
+## Loading Instructions
+
+```python
+from datasets import load_dataset
+
+ds = load_dataset("YOUR_ORG/MMPD-Bench")
+print(ds)
+print(ds["train"][0].keys())
+```
+
+```python
+import torch
+
+row = ds["train"][0]
+mueller = torch.tensor(row["mueller"], dtype=torch.float32)
+target = torch.tensor(row["target"], dtype=torch.float32)
+```
+
+## Ethics and Limitations
+
+TODO: Add de-identification statement, biological ground-truth limitation, and
+recommended use boundaries.
+
+## License
+
+TODO: Add final dataset license and any access restrictions.
+
+## Citation
+
+TODO: Add paper citation and BibTeX.
+
+## Contact
+
+TODO: Add maintainer contact details.