All functionalities working fine

app.py

@@ -0,0 +1,29 @@
+import gradio as gr
+import os
+import numpy as np
+from onnx_inference import emotions_detector
+
+class_names = ['angry', 'happy', 'sad']
+
+
+def predict(img):
+    img = np.array(img)
+    onnx_pred, time_taken = emotions_detector(img)
+
+    pred_labels_and_probs = {class_names[i]: float(
+        onnx_pred[0][0][i]) for i in range(len(class_names))}
+
+    return pred_labels_and_probs, time_taken
+
+
+title = "Human Emotion Detection 😭🤣🥹"
+description = "An EfficientNet ONNX quantized feature extractor computer vision model to classify images and detect the emotion of the person in it.(Uploaded image should be of a single person)"
+article = "Full Source code from scratch can be found in the huggingface Space...."
+
+# Create examples list from "examples/" directory
+example_list = [["examples/" + example] for example in os.listdir("examples")]
+
+demo = gr.Interface(fn=predict, inputs=gr.Image(type='pil'), outputs=[gr.Label(num_top_classes=3, label='Predictions'), gr.Number(
+    label="Prediction time (s)")], examples=example_list, title=title, description=description, article=article)
+
+demo.launch()

detect.py

@@ -0,0 +1,10 @@
+from PIL import Image
+from io import BytesIO
+import numpy as np
+
+
+def emo_router(im):
+    print(f"the Image: {im}")
+    image = np.array(im)
+
+    return image

eff_quantized.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c278206e78d48fc8ab5151bc22e2636faad7bb41323ac5f0b6bde72079ebf72
+size 63147650

onnx_inference.py

@@ -0,0 +1,27 @@
+import onnxruntime as rt
+import cv2
+import numpy as np
+import time
+
+providers = ['CPUExecutionProvider']
+m_q = rt.InferenceSession(
+    "eff_quantized.onnx", providers=providers)
+
+
+def emotions_detector(img_array):
+    time_init = time.time()
+
+    # Check if image is in grayscale and convert to rgb
+    if len(img_array.shape) == 2:
+        img_array = cv2.cvtColor(img_array, cv2.COLOR_GRAY2RGB)
+
+    # resize layer
+    test_image = cv2.resize(img_array, (256, 256))
+    im = np.float32(test_image)
+    img_array = np.expand_dims(im, axis=0)
+
+    onnx_pred = m_q.run(['dense_2'], {"input_1": img_array})
+
+    time_elapsed = time.time() - time_init
+
+    return onnx_pred, time_elapsed