dccc2025
/

TeX-UNet

@@ -87,6 +87,25 @@ This repository releases two inference checkpoints from
 [TeX-1500: A Paired Real-World LWIR Hyperspectral Dataset and Benchmark for
 Temperature-Emissivity-Texture Decomposition](https://arxiv.org/abs/2606.03806).
 ## Files
 ```text
@@ -129,6 +148,18 @@ TeX-UNet/
 Both checkpoints are stored as `safetensors` and contain only the model state
 dict, not optimizer state.
 ## Benchmark
 The initial benchmark reports TeX-UNet inversion results on held-out DARPA IH
@@ -204,8 +235,50 @@ python scripts/verify_load.py \
 The dataset is hosted separately at
 [jialelin2007/TeX-1500](https://huggingface.co/datasets/jialelin2007/TeX-1500).
-It is a gated dataset; log in to Hugging Face and accept the dataset access
-conditions before downloading files:
 ```bash
 hf auth login
@@ -217,12 +290,12 @@ hf download jialelin2007/TeX-1500 \
 The dataset license and access terms are governed by the dataset repository.
-## Links
-- Paper: https://arxiv.org/abs/2606.03806
-- Dataset: https://huggingface.co/datasets/jialelin2007/TeX-1500
-- Code: https://github.com/dccc2025/TeX-1500
-- Model: https://huggingface.co/dccc2025/TeX-UNet
 ## Limitations
@@ -237,10 +310,13 @@ The dataset license and access terms are governed by the dataset repository.
 ## Citation
 ```bibtex
-@article{dai2026tex1500,
-  title = {TeX-1500: A Paired Real-World LWIR Hyperspectral Dataset and Benchmark for Temperature-Emissivity-Texture Decomposition},
-  author = {Dai, Cheng and Lin, Jiale and Xu, Hongyi and Song, Bingxuan and Xie, Ziyang and Bao, Fanglin},
-  journal = {arXiv preprint arXiv:2606.03806},
-  year = {2026}
 }
 ```

 [TeX-1500: A Paired Real-World LWIR Hyperspectral Dataset and Benchmark for
 Temperature-Emissivity-Texture Decomposition](https://arxiv.org/abs/2606.03806).
+## Authors
+- Cheng Dai ([Hugging Face](https://huggingface.co/dccc2025), [GitHub](https://github.com/dccc2025))*
+- Jiale Lin ([Hugging Face](https://huggingface.co/jialelin2007), [GitHub](https://github.com/jialelin2007))*
+- Hongyi Xu
+- Bingxuan Song
+- Ziyang Xie
+- Fanglin Bao
+*Equal Contribution.
+## Links
+- Hugging Face Papers: [TeX-1500](https://huggingface.co/papers/2606.03806)
+- arXiv: [2606.03806](https://arxiv.org/abs/2606.03806)
+- Dataset: [jialelin2007/TeX-1500](https://huggingface.co/datasets/jialelin2007/TeX-1500)
+- Code: [dccc2025/TeX-1500](https://github.com/dccc2025/TeX-1500)
+- Model: [dccc2025/TeX-UNet](https://huggingface.co/dccc2025/TeX-UNet)
 ## Files
 ```text
 Both checkpoints are stored as `safetensors` and contain only the model state
 dict, not optimizer state.
+Each checkpoint uses the same TeX-UNet-v2 architecture:
+- input HSI tensor: `[B, 64, H, W]`
+- input wavelength tensor: `[64]` or `[B, 64]`, in micrometers
+- outputs:
+  - `T`: `[B, 1, H, W]`, temperature
+  - `e`: `[B, 64, H, W]`, spectral emissivity
+  - `X`: `[B, 1, H, W]`, texture
+- wavelength range: 6-14 micrometers
+- inference tiling: `224 x 224` patches, stride `112`, raised-cosine blending
+- released parameter count: `34,279,155`
 ## Benchmark
 The initial benchmark reports TeX-UNet inversion results on held-out DARPA IH
 The dataset is hosted separately at
 [jialelin2007/TeX-1500](https://huggingface.co/datasets/jialelin2007/TeX-1500).
+The local dataset release notes describe TeX-1500 as a paired LWIR HSI benchmark
+with 1,522 calibrated real-scene HSI-TeX pairs from DARPA pushbroom data and
+FTIR acquisitions.
+The dataset repository is separate from this model repository. The current
+dataset release is a preview containing one example sample for format inspection
+and visualization; the full dataset is planned for a later release.
+Current preview layout:
+```text
+data/sample_0001/
+  hsi.mat
+  hsi_noisy.mat
+  T.mat
+  e.mat
+  X.mat
+  previews/
+    T.png
+    e.png
+    X.png
+    hsi_band.png
+```
+The preview `.mat` files use MATLAB v5 format:
+| File | Variable | Shape | Dtype | Description |
+|---|---|---:|---|---|
+| `hsi.mat` | `denoised_hsi_original` | `[260, 1500, 256]` | `single` | Denoised calibrated LWIR HSI cube. |
+| `hsi.mat` | `hsi_wav` | `[1, 256]` | `double` | HSI wavelength grid. |
+| `hsi.mat` | `good_band_indices` | `[1, 230]` | `int64` | 0-indexed valid band indices used by the processing pipeline. |
+| `hsi.mat` | `calibrated_sky` | `[1, 256]` | `double` | Calibrated sky signal estimate. |
+| `hsi.mat` | `observed_sky` | `[1, 256]` | `double` | Observed sky signal estimate. |
+| `hsi.mat` | `working_wav` | `[1, 256]` | `single` | Calibrated working wavelength grid. |
+| `hsi_noisy.mat` | `hsi_noisy` | `[260, 1500, 256]` | `single` | Noisy input HSI cube. |
+| `T.mat` | `T` | `[260, 1500]` | `single` | Temperature field. |
+| `e.mat` | `e` | `[260, 1500, 256]` | `single` | Spectral emissivity field. |
+| `X.mat` | `X` | `[260, 1500]` | `single` | Texture field. |
+Preview PNG files are for visual inspection only and should not be treated as
+numeric labels.
+If the Hugging Face dataset repository requires authentication, log in and
+accept its access terms before downloading files:
 ```bash
 hf auth login
 The dataset license and access terms are governed by the dataset repository.
+## License
+This model repository is released under Apache-2.0. The TeX-1500 dataset files
+are hosted separately and are distributed under the PolyForm Noncommercial
+License 1.0.0 in the dataset repository. TeX-1500 dataset files are not included
+in this model repository.
 ## Limitations
 ## Citation
 ```bibtex
+@misc{dai2026tex1500pairedrealworldlwir,
+      title={TeX-1500: A Paired Real-World LWIR Hyperspectral Dataset and Benchmark for Temperature-Emissivity-Texture Decomposition},
+      author={Cheng Dai and Jiale Lin and Hongyi Xu and Bingxuan Song and Ziyang Xie and Fanglin Bao},
+      year={2026},
+      eprint={2606.03806},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV},
+      url={https://arxiv.org/abs/2606.03806},
 }
 ```