init hf repo

app.py

@@ -0,0 +1,20 @@
+import gradio as gr
+
+def process_image(input_image):
+    # Your ML model code goes here. For example:
+    # annotated_image = annotate_image(input_image)
+    # confidence = get_classification_confidence(input_image)
+    annotated_image = None
+    confidence = 0
+    return annotated_image, f"Classification Confidence: {confidence}%"
+
+# Define the Gradio interface
+
+if __name__ == '__main__':
+    iface = gr.Interface(
+        fn=process_image,
+        inputs="image",
+        outputs=["image", "text"],
+        title="AMBULANT Benthic Identification System"
+    )
+    iface.launch(auth=("admin", "pass1234"), share=True)

data_analyser.py

@@ -0,0 +1,20 @@
+from inferencing import process_image
+from habitat_classification import classify_habitat
+
+
+def process_images(images):
+    processed_images = []
+    for image in images:
+        processed_images.push(process_image(image.read()))
+    return processed_images
+
+
+def classify_habitat(species_arr):
+    result = classify_habitat(species_arr)
+    result_str = []
+    result_str.push(f"Predicted Classification: {result[0]}")
+    result_str.push(f"Confidence: {result[1]:.2f}")
+    result_str.push("Classification Probabilities:")
+    for classification, probability in result[2].items():
+        result_str.push(f"{classification}: {probability:.2f}")
+    return result_str

habitat_classification.py

@@ -0,0 +1,64 @@
+import pandas as pd
+
+# Step 1: Read the data
+species_data = pd.read_csv('species_data.csv')
+habitat_data = pd.read_csv('habitat_data.csv')
+
+# Normalize the column names
+common_column_names = [
+    "Species",
+    "Habitat: Hard beds",
+    "Habitat: Rocky bottoms",
+    "Habitat: Seagrass meadows",
+    "Habitat: Coralligenous bottoms",
+    "Habitat: Detritic bottoms",
+    "Habitat: Sandy bottoms",
+    "Habitat: Muddy bottoms",
+    "Habitat: Other organisms"
+]
+
+species_data.columns = common_column_names
+habitat_data.columns = common_column_names[1:] + ["Final classification"]
+
+# Step 2: Convert "Yes"/"No" to 1/0 and normalize the values to calculate weights
+species_data.replace({"Yes": 1, "No": 0}, inplace=True)
+habitat_data.replace({"Yes": 1, "No": 0}, inplace=True)
+
+species_weights = species_data.iloc[:, 1:].div(species_data.iloc[:, 1:].sum(axis=1), axis=0)
+habitat_weights = habitat_data.iloc[:, :-1].div(habitat_data.iloc[:, :-1].sum(axis=1), axis=0)
+
+# Step 3: Create dictionaries to map species to habitats and habitats to final classifications with weights
+species_to_habitat_weights = species_data.set_index('Species').T.to_dict('series')
+final_classification_series = habitat_data['Final classification']
+#habitat_data.drop(columns=['Final classification'], inplace=True)
+
+habitat_to_classification_weights = habitat_data.set_index(final_classification_series).T.to_dict('series')
+
+# Step 4: Define the classification function
+def classify_habitat(species_list):
+    habitat_votes = {col: 0 for col in species_data.columns[1:]}
+    classification_votes = {row['Final classification']: 0 for _, row in habitat_data.iterrows()}
+
+    # Let each species vote for habitats
+    for species in species_list:
+        if species in species_to_habitat_weights:
+            for i, habitat in enumerate(habitat_votes.keys()):
+                habitat_votes[habitat] += species_to_habitat_weights[species][i]
+
+    # Let habitats vote for the final classification
+    for habitat, votes in habitat_votes.items():
+        for classification in classification_votes.keys():
+            classification_votes[classification] += habitat_to_classification_weights[classification][habitat] * votes
+
+
+    # Get the classification with the highest vote
+    predicted_classification = max(classification_votes, key=classification_votes.get)
+
+    # Calculate confidence
+    total_votes = sum(classification_votes.values())
+    confidence = classification_votes[predicted_classification] / total_votes
+
+    # Calculate individual classification probabilities
+    classification_probabilities = {key: value / total_votes for key, value in classification_votes.items()}
+
+    return predicted_classification, confidence, classification_probabilities

habitat_data.csv

@@ -0,0 +1,15 @@
+Habitat: Hard beds,Habitat: Rocky bottoms,Habitat: Seagrass meadows,Habitat: Coralligenous bottoms,Habitat: Detritic bottoms,Habitat: Sandy bottoms,Habitat: Muddy bottoms,Habitat: Other organisms,Final classification
+Yes,Yes,Yes,Yes,No,No,No,Yes,MB151 Biocenosis of Mediterranean infralittoral algae
+Yes,Yes,Yes,Yes,No,No,No,Yes,MB251 Mediterannean biogenic reef assemblages of the infralittoral algae biocenosis
+Yes,Yes,Yes,Yes,No,Yes,No,Yes,MB252 Biocenosis of Posidonia oceanica
+No,Yes,Yes,No,Yes,Yes,No,Yes,MB351 Biocenosis of Mediterranean coarse sands and fine gravels mixed by the waves
+No,Yes,Yes,No,Yes,Yes,No,Yes,MB352 Biocenosis of Mediterranean coarse sands and fine gravels under the influence of bottoms currents
+No,Yes,Yes,No,Yes,No,No,Yes,MB353 Biocenosis of Mediterranean infralittoral pebbles
+No,No,Yes,No,No,Yes,No,Yes,MB551 Biocenosis of Mediterranean fine surface sands
+No,No,Yes,No,No,Yes,No,Yes,MB552 Biocenosis of Mediterranean well sorted fine sands 
+No,No,Yes,No,No,Yes,Yes,Yes,MB553 Biocenosis of Mediterranean superficial muddy sands in sheltered waters
+Yes,Yes,Yes,Yes,No,No,No,Yes,MC151 Coralligenous biocenosis
+Yes,Yes,Yes,Yes,No,No,No,Yes,MC251 Coralligenous platforms
+No,Yes,Yes,No,Yes,Yes,No,Yes,MC351 Biocenosis of Mediterranean coastal detritic bottoms (without rodolithes)
+No,Yes,Yes,Yes,Yes,Yes,No,Yes,MC352 Assemblages of Mediterranean coastal detritic bottoms biocenosis with rhodolithes
+No,No,Yes,No,Yes,Yes,No,Yes,MC551 Biocenosis of Mediterranean coarse sands and fine gravels under the influence of bottom currents

inferencing.py

@@ -0,0 +1,44 @@
+from models.yolo8.ultralytics.ultralytics import YOLO
+import pandas as pd
+import tempfile
+import torch
+
+
+# Function to display detections and segmentation masks
+def display_detections(frame, results):
+    detections = []
+    for result in results:
+        if result.boxes.data.nelement() != 0:
+            for detection in result.boxes.data:
+                x1, y1, x2, y2, conf, cls = detection.cpu().numpy()
+                class_name = result.names[int(cls)]
+                detections.append([class_name, conf, x1, y1, x2, y2])
+            outframe = results[0].plot()
+            return outframe, detections
+
+    print("No detections")
+    return frame, detections
+
+
+def process_image(input_image):
+    results = model(input_image)
+    processed_frame, detections = display_detections(input_image, results)
+
+    if not detections:  # Check if detections list is empty
+        return processed_frame, "No detections found"
+
+    # Convert detections to a DataFrame and then save to a temporary CSV file
+    df = pd.DataFrame(detections, columns=["Class", "Confidence", "X1", "Y1", "X2", "Y2"])
+    temp_file = tempfile.NamedTemporaryFile(delete=False, suffix='.csv')
+    df.to_csv(temp_file.name, index=False)
+
+    return processed_frame, temp_file.name
+
+
+# Load YOLO model
+model = YOLO('../models/segmentation/latest.pt')
+model.conf = 0.2
+
+if torch.backends.cuda.is_built():
+    print(f"CUDA available: {torch.cuda.is_available()}")
+    print(f"Device name: {torch.cuda.get_device_name(0)}")

species_data.csv

@@ -0,0 +1,18 @@
+Species,Habitat: Hard beds,Habitat: Rocky bottoms,Habitat: Seagrass meadows,Habitat: Coralligenous bottoms,Habitat: Detritic bottoms,Habitat: Sandy bottoms,Habitat: Muddy bottoms,Habitat: Other organisms
+Asparagopsis taxiformis,Yes,Yes,No,No,No,No,No,No
+Liagora viscida,No,Yes,No,No,No,No,No,No
+Lophocladia Lallemandii,No,No,Yes,No,No,No,No,Yes
+Codium bursa,Yes,No,No,No,Yes,No,No,No
+Ulva compressa,Yes,Yes,Yes,Yes,Yes,Yes,Yes,Yes
+Dictyopteris polypodioides,Yes,Yes,No,No,No,No,No,No
+Dictyota dichotoma,Yes,Yes,No,No,No,No,No,No
+Padina pavonica,Yes,Yes,No,Yes,No,No,No,No
+Sargassum vulgare,Yes,Yes,Yes,Yes,Yes,No,No,No
+Cymodocea nodosa,No,No,No,No,No,Yes,No,No
+Halophila stipulacea,No,No,No,No,No,Yes,Yes,No
+Posidonia oceanica,Yes,Yes,No,No,Yes,Yes,Yes,No
+Centrostephanus longispinus,Yes,Yes,No,No,No,No,No,No
+Chondrilla nucula,Yes,Yes,Yes,No,No,No,No,No
+Sarcotragus spinosulus,Yes,Yes,No,No,No,No,No,No
+Scalarispongia scalaris,Yes,Yes,Yes,Yes,Yes,No,No,No
+Hermodice carunculata,Yes,Yes,No,No,Yes,No,No,No