Upload folder using huggingface_hub

README.md

@@ -1,12 +1,6 @@
 ---
-title: Ocsai Drawings
-emoji: 🚀
-colorFrom: green
-colorTo: blue
+title: Ocsai_Drawings
+app_file: gradio-ocsai-d.py
 sdk: gradio
 sdk_version: 4.27.0
-app_file: app.py
-pinned: false
 ---
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

gradio-ocsai-d.py

@@ -0,0 +1,113 @@
+from pathlib import Path
+from transformers import pipeline
+from PIL import Image
+import gradio as gr
+import pandas as pd
+import numpy as np
+import altair as alt
+
+
+prompt_images = {
+    "Images_11": "blanks/Images_11_blank.png",
+    "Images_4": "blanks/Images_4_blank.png",
+    "Images_17": "blanks/Images_17_blank.png",
+    "Images_9": "blanks/Images_9_blank.png",
+    "Images_3": "blanks/Images_3_blank.png",
+    "Images_8": "blanks/Images_8_blank.png",
+    "Images_13": "blanks/Images_13_blank.png",
+    "Images_15": "blanks/Images_15_blank.png",
+    "Images_12": "blanks/Images_12_blank.png",
+    "Images_7": "blanks/Images_7_blank.png",
+    "Images_56": "blanks/Images_56_blank.png",
+    "Images_19": "blanks/Images_19_blank.png",
+}
+
+dist = pd.read_csv('./score_norm_distribution.csv', dtype=float)
+
+
+def get_percentile(score):
+    return dist[dist['score_norm'] <= score].iloc[-1, 0]
+
+
+def inverse_scale(logits):
+    # undo the min-max scaling that was done from the JRT range to 0-1
+    scaler_params = {'min': -3.024, 'max': 3.164, 'range': 6.188}
+    return logits * (scaler_params['range']) + scaler_params['min']
+
+
+def get_predictions(img_dict):
+    # gradio passes a dictionary with background, composite, and layers
+    # the composite is what we want
+    img = img_dict['composite']
+    predictions = classifier(img)
+    return {
+        'originality' : np.round(predictions[0]['score'], 2),
+        'jrt' : np.round(inverse_scale(0), 2),
+        'percentile': get_percentile(predictions[0]['score'])
+    }
+
+
+base_chart = alt.Chart(dist).mark_line().encode(
+    x='percentile',
+    y='score_norm'
+)
+
+def classify_image(img_dict):
+    # gradio passes a dictionary with background, composite, and layers
+    # the composite is what we want
+    img = img_dict['composite']
+    p = get_predictions(img_dict)
+
+    percentile_mark = alt.Chart(pd.DataFrame({'y': [p['originality']]})).mark_rule(color='red').encode(y='y')
+
+    # Text annotation for the percentile mark
+    text = alt.Chart(pd.DataFrame({'y': [p['originality']], 'text': [f"Percentile: {p['percentile']}; Normalized Score: {p['originality']}"]})).mark_text(
+        align='left',
+        baseline='middle',
+        dx=7, dy=-8  # Nudges text to right so it doesn't overlap with the line
+    ).encode(
+        y='y',
+        text='text'
+    )
+    return base_chart + percentile_mark + text
+
+
+def update_editor(background, img_editor):
+    # Clear layers and set the selected background
+    img_editor['background'] = background
+    img_editor['layers'] = []
+    img_editor['composite'] = None
+    return img_editor
+
+
+classifier = pipeline("image-classification", model='POrg/ocsai-d-large')
+
+editor = gr.ImageEditor(type='pil',
+                        value=dict(
+                            background=Image.open(prompt_images['Images_11']),
+                            composite=None,
+                            layers=[]
+                        ),
+                        brush=gr.Brush(
+                            default_size=2,
+                            colors=["#000000", '#333333', '#666666'],
+                            color_mode="fixed"
+                        ),
+                        transforms=[],
+                        sources=('upload', 'clipboard'),
+                        layers=False
+                        )
+
+examples = []
+for k, v in prompt_images.items():
+    examples.append(dict(background=Image.open(v), composite=None, layers=[]))
+
+demo = gr.Interface(fn=classify_image,
+                     inputs=[editor],
+                     outputs=gr.Plot(),
+                     title="Ocsai-D",
+                     description="Complete the drawing and classify the originality. Examples are from MTCI ([Barbot 2018](https://pubmed.ncbi.nlm.nih.gov/30618952/)). Choose the brush icon below the image to start editing.",
+                     examples=examples
+                     )
+
+demo.launch(debug=True)

requirements.txt

@@ -0,0 +1,5 @@
+altair
+transformers
+pillow
+pandas
+numpy

score_norm_distribution.csv

@@ -0,0 +1,102 @@
+percentile,score_norm
+0,0.0
+1,0.2658954269536899
+2,0.29457921333383075
+3,0.30310490366607795
+4,0.3112277375239338
+5,0.3200521694123999
+6,0.3281155279694775
+7,0.33505868144658657
+8,0.3392089731525685
+9,0.34112156925606424
+10,0.3469171398619739
+11,0.3511710111618856
+12,0.3568430068136271
+13,0.36180990478859454
+14,0.3620405096808955
+15,0.3660549175130392
+16,0.3713281681316126
+17,0.3758821435250105
+18,0.38007216818894674
+19,0.38286859218641567
+20,0.3851654423305501
+21,0.38878772721364474
+22,0.3933188590264441
+23,0.3985148402511076
+24,0.4020135391925982
+25,0.4059742928167005
+26,0.4116649490408034
+27,0.4158741294366881
+28,0.42095292315008237
+29,0.42500796302503663
+30,0.42947442498133825
+31,0.43421050644034465
+32,0.43848988145355117
+33,0.4428767461965395
+34,0.44774875322791224
+35,0.4520888259135842
+36,0.4561818524913554
+37,0.45951053385996776
+38,0.4654170913642474
+39,0.46936765339064945
+40,0.47374096009598826
+41,0.4778733813156729
+42,0.48168420678878365
+43,0.4858078800434697
+44,0.489977885665859
+45,0.49222064341619076
+46,0.4935026730860239
+47,0.49741882057510833
+48,0.5018924541619932
+49,0.5059561160864396
+50,0.5104564475386445
+51,0.5147230242331543
+52,0.5183986906290778
+53,0.5225670918406934
+54,0.5259651360191281
+55,0.5288777252975223
+56,0.5324189434087236
+57,0.5366995325233963
+58,0.5407957310743421
+59,0.5445732698714345
+60,0.5486969650375954
+61,0.5517929075887892
+62,0.5552200888240536
+63,0.5591196062209663
+64,0.5634000324385277
+65,0.5671347562440855
+66,0.5706120211744352
+67,0.573789230587204
+68,0.577790110565303
+69,0.5822551850499115
+70,0.586409034837951
+71,0.590284528513099
+72,0.5940106484952778
+73,0.5977210767573214
+74,0.602416228103494
+75,0.6066765752140205
+76,0.6120308682527665
+77,0.6158761623245004
+78,0.6200640057567599
+79,0.6246296981878875
+80,0.6293587808987731
+81,0.6344289521636639
+82,0.6395740140855798
+83,0.6448253685442593
+84,0.650423636218271
+85,0.6558870284254413
+86,0.6615215893114951
+87,0.6666412816994163
+88,0.6733659429202951
+89,0.6803897908526316
+90,0.6873526726184538
+91,0.6954822613275397
+92,0.7039240961497323
+93,0.7140753120540401
+94,0.7245636076473253
+95,0.7356199857393928
+96,0.7505542818394643
+97,0.7701851740360041
+98,0.794328590193434
+99,0.8426026407211062
+100,0.9929171993962822