app.py

import gradio as gr
import torch
from io import BytesIO
import os

import cv2
import gradio as gr
import numpy as np
import requests
from PIL import Image

import gradio as gr
import cv2
import tempfile
import numpy as np
import torch
from torchvision import transforms
from PIL import Image
import matplotlib.pyplot as plt
from io import BytesIO

# Load the YOLO model
from models.common import DetectMultiBackend



weights_path = "./last.pt"
device = torch.device("cpu")  # Correctly define the device
model = DetectMultiBackend(weights_path, device=device)  # Load YOLOv5 model correctly
model.eval()

# model_path = "./last.pt"
# model = torch.jit.load(model_path, map_location=torch.device("cpu"))
# model.eval()
# transform=transforms.Compose([
#     transforms.ToPILImage(),
#     transforms.Resize((512,640)),
#     transforms.ToTensor()
# ])
transform = transforms.Compose([
    transforms.ToPILImage(),  # Ensure input is a PIL image
    transforms.Resize((512, 640)),
    transforms.ToTensor()
])
# transform = transforms.Compose([
#     transforms.Resize((640, 640)),
#     transforms.ToTensor(),
# ])

OBJECT_NAMES = ['enemies']

def detect_objects_in_image(image):
       """
    Detect objects in the given image.
    """
    # Ensure image is a PIL Image
    if isinstance(image, torch.Tensor):
        image = transforms.ToPILImage()(image)  # Convert tensor to PIL image
    
    if isinstance(image, Image.Image):
        orig_w, orig_h = image.size  # PIL image size returns (width, height)
    else:
        raise TypeError(f"Expected a PIL Image but got {type(image)}")

    # Apply transformation
    img_tensor = transform(image).unsqueeze(0)



    with torch.no_grad():
        pred = model(img_tensor)[0]

    if isinstance(pred[0], torch.Tensor):
        pred = [p.cpu().numpy() for p in pred]

    pred = np.concatenate(pred, axis=0)
    conf_thres = 0.25
    mask = pred[:, 4] > conf_thres
    pred = pred[mask]

    if len(pred) == 0:
        return Image.fromarray(np.array(image)), None  # Return only image and None for graph

    boxes, scores, class_probs = pred[:, :4], pred[:, 4], pred[:, 5:]
    class_ids = np.argmax(class_probs, axis=1)

    boxes[:, 0] = boxes[:, 0] - (boxes[:, 2] / 2)
    boxes[:, 1] = boxes[:, 1] - (boxes[:, 3] / 2)
    boxes[:, 2] = boxes[:, 0] + boxes[:, 2]
    boxes[:, 3] = boxes[:, 1] + boxes[:, 3]

    boxes[:, [0, 2]] *= orig_w / 640
    boxes[:, [1, 3]] *= orig_h / 640
    boxes = np.clip(boxes, 0, [orig_w, orig_h, orig_w, orig_h])

    indices = cv2.dnn.NMSBoxes(boxes.tolist(), scores.tolist(), conf_thres, 0.5)

    object_counts = {name: 0 for name in OBJECT_NAMES}
    img_array = np.array(image)

    if len(indices) > 0:
        for i in indices.flatten():
            x1, y1, x2, y2 = map(int, boxes[i])
            cls = class_ids[i]
            object_name = OBJECT_NAMES[cls] if cls < len(OBJECT_NAMES) else f"Unknown ({cls})"
            if object_name in object_counts:
                object_counts[object_name] += 1
            cv2.rectangle(img_array, (x1, y1), (x2, y2), (0, 255, 0), 2)
            cv2.putText(img_array, f"{object_name}: {scores[i]:.2f}", (x1, y1 - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)

    # Generate and return graph instead of dictionary
    graph_image = generate_vehicle_count_graph(object_counts)

    return Image.fromarray(img_array), graph_image  # Now returning only 2 outputs


# def generate_vehicle_count_graph(object_counts):
#     color_palette = ['#4C9ACD', '#88B8A3', '#7F9C9C', '#D1A3B5', '#A1C6EA', '#FFB6C1', '#F0E68C', '#D3B0D8', '#F8A5D1', '#B8B8D1']
    
#     fig, ax = plt.subplots(figsize=(8, 5))
#     labels = list(object_counts.keys())
#     values = list(object_counts.values())
    
#     ax.bar(labels, values, color=color_palette[:len(labels)])
    
#     ax.set_xlabel("Vehicle Categories", fontsize=12, fontweight='bold')
#     ax.set_ylabel("Number of Vehicles", fontsize=12, fontweight='bold')
#     ax.set_title("Detected Vehicles in Image", fontsize=14, fontweight='bold')
    
#     plt.xticks(rotation=45, ha='right', fontsize=10)
#     plt.yticks(fontsize=10)
    
#     plt.tight_layout()

#     buf = BytesIO()
#     plt.savefig(buf, format='png')
#     buf.seek(0)
    
#     return Image.open(buf)

def generate_vehicle_count_graph(object_counts):
    color_palette = ['#4C9ACD', '#88B8A3', '#7F9C9C', '#D1A3B5', '#A1C6EA', '#FFB6C1', '#F0E68C', '#D3B0D8', '#F8A5D1', '#B8B8D1']
    
    fig, ax = plt.subplots(figsize=(8, 5))
    labels = list(object_counts.keys())
    values = list(object_counts.values())
    
    ax.bar(labels, values, color=color_palette[:len(labels)])
    ax.set_xlabel("Vehicle Categories", fontsize=12, fontweight='bold')
    ax.set_ylabel("Number of Vehicles", fontsize=12, fontweight='bold')
    ax.set_title("Detected Vehicles in Image", fontsize=14, fontweight='bold')
    
    plt.xticks(rotation=45, ha='right', fontsize=10)
    plt.yticks(fontsize=10)
    plt.tight_layout()
    
    buf = BytesIO()
    plt.savefig(buf, format='png')
    buf.seek(0)
    
    plt.close(fig)  # ✅ CLOSE THE FIGURE TO FREE MEMORY
    
    return Image.open(buf)
    
def detect_objects_in_video(video_input):
    cap = cv2.VideoCapture(video_input)
    if not cap.isOpened():
        return "Error: Cannot open video file.", None  # Returning a second value (None) to match expected outputs

    frame_width, frame_height, fps = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH)), int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT)), int(cap.get(cv2.CAP_PROP_FPS))
    temp_video_output = tempfile.NamedTemporaryFile(suffix=".mp4", delete=False).name
    out = cv2.VideoWriter(temp_video_output, cv2.VideoWriter_fourcc(*'mp4v'), fps, (frame_width, frame_height))

    # Initialize the counts for vehicle categories
    total_counts = {name: 0 for name in ['car', 'truck', 'bus', 'motorcycle', 'bicycle']}

    while cap.isOpened():
        ret, frame = cap.read()
        if not ret:
            break

        image = Image.fromarray(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB))

        # Get frame with detected objects and graph
        frame_with_boxes, graph_image = detect_objects_in_image(image)

        # Convert image back to OpenCV format for writing video
        out.write(cv2.cvtColor(np.array(frame_with_boxes), cv2.COLOR_RGB2BGR))

    cap.release()
    out.release()

    return temp_video_output, graph_image  # Return both expected outputs

def greet(name):
    return "Hello " + name + "!!"

demo = gr.Interface(fn=greet, inputs="text", outputs="text")

from urllib.request import urlretrieve

# get image examples from github
urlretrieve("https://github.com/SamDaaLamb/ValorantTracker/blob/main/clip2_-1450-_jpg.jpg?raw=true", "clip2_-1450-_jpg.jpg") # make sure to use "copy image address when copying image from Github"
urlretrieve("https://github.com/SamDaaLamb/ValorantTracker/blob/main/clip2_-539-_jpg.jpg?raw=true", "clip2_-539-_jpg.jpg")
examples = [ # need to manually delete cache everytime new examples are added
    ["clip2_-1450-_jpg.jpg"], 
    ["clip2_-539-_jpg.jpg"]]


# define app features and run
title = "SpecLab Demo"
description = "<p style='text-align: center'>Gradio demo for an ASPP model architecture trained on the SpecLab dataset. To use it, simply add your image, or click one of the examples to load them. Since this demo is run on CPU only, please allow additional time for processing. </p>"
article = "<p style='text-align: center'><a href='https://github.com/Nano1337/SpecLab'>Github Repo</a></p>"
css = "#0 {object-fit: contain;} #1 {object-fit: contain;}"
demo = gr.Interface(fn=detect_objects_in_image, 
                    title=title, 
                    description=description,
                    article=article,
                    inputs=gr.Image(elem_id=0, show_label=False), 
                    outputs=gr.Image(elem_id=1, show_label=False),
                    css=css, 
                    examples=examples, 
                    cache_examples=True,
                    allow_flagging='never')
demo.launch()