concordia: sync data card and pipeline grid parameters with GitHub

Syncs the concordia (SynthUS-FSA) card and pipeline on the Hub with the reviewed state on GitHub (open-h/OpenH-RF).

`pipeline.yaml` gains the `parameters` block the reference images were made with: a 45 x 40 mm field of view starting 10 mm from the transducer face, a 610 x 542 grid so pixels are isotropic at ~74 um (the stored `data/image` grid is 542 x 384 and anisotropic, so rendering it as square pixels stretches the phantom vertically), `f_number: 1.75` and a -60 dB dynamic range. Without these, zea falls back to an aperture-wide, full-depth default grid at `f_number: 1.0`, which does not reproduce the card's figures.

Adds the figures the card shows: `assets/classes.png`, one B-mode per phantom class, and `assets/reference_capture.png`, the per-capture validation figure that overlays label, reconstruction and scatterer cloud on shared mm axes. The card previously pointed at an `examples/` folder that is not published; it now points at `assets/`. `assets/main.png` is the release image and is not embedded in the card.

The card also gains a title, the class montage, and the `zea process` one-liner that renders `data/image_0005.hdf5` straight from the Hub.

Files changed: 5.

```
+ concordia/assets/classes.png
+ concordia/assets/main.png
+ concordia/assets/reference_capture.png
~ concordia/README.md
~ concordia/pipeline.yaml
```

concordia/README.md

@@ -16,6 +16,24 @@ size_categories:
   - 1K<n<10K
 ---
 
+# SynthUS-FSA
+
+![B-mode reconstructions of the five SynthUS-FSA phantom classes](assets/classes.png)
+
+The five phantom classes, each reconstructed from `data/raw_data`:
+`image_0005`, `image_0470`, `image_0640`, `image_1100`, `image_1808`.
+
+`zea` renders any capture straight from the Hub with the
+`pipeline.yaml` in this folder. Try it out with the following command:
+
+```bash
+zea process \
+  --dataset hf://nvidia/OpenH-RF/concordia/data/image_0005.hdf5 \
+  --config hf://nvidia/OpenH-RF/concordia/pipeline.yaml \
+  --n-frames 1 \
+  --save-as png
+```
+
 ## Dataset Description
 
 SynthUS-FSA is a fully synthetic corpus of pre-beamformed full-synthetic-aperture (FSA)
@@ -259,8 +277,7 @@ reconstruction on physical mm axes next to the capture's class-specific label
 `data/diverse_source_image` for diverse, none for point-target — and the
 `data/scatterers` cloud coloured by |amplitude|, all on shared equal-aspect mm
 axes, confirming the label, reconstruction, and scatterer field are spatially
-registered. `examples/` holds one such figure for a representative capture of
-each of the five classes.
+registered. [`assets/reference_capture.png`](assets/reference_capture.png) is one such figure.
 
 ## Known Issues
 

concordia/pipeline.yaml

@@ -1,3 +1,24 @@
+parameters:
+    # Imaging FOV: 45 mm lateral x 40 mm axial, starting 10 mm from the
+    # transducer face. This is a reconstruction-time choice, not part of the zea
+    # scan spec, so without it zea falls back to an aperture-wide, full-depth
+    # default grid.
+    xlims:
+    - -0.0225
+    - 0.0225
+    zlims:
+    - 0.01
+    - 0.05
+    # Isotropic pixels (~74 um both ways). data/image is stored on a 542 x 384
+    # anisotropic grid; rendering that grid as square pixels stretches the
+    # phantom vertically.
+    grid_size_x: 610
+    grid_size_z: 542
+    # zea's default is 1.0, which is not what produced the reference images.
+    f_number: 1.75
+    dynamic_range:
+    - -60
+    - 0
 pipeline:
     operations:
     -   name: keras.ops.cast