Spaces:
Sleeping
Sleeping
File size: 3,685 Bytes
53f723c 664f840 53f723c cf2e328 664f840 53f723c f4e9d9d cf2e328 f4e9d9d cf2e328 9d4eaf2 cf2e328 f4e9d9d cf2e328 f4e9d9d cf2e328 f4e9d9d cf2e328 f4e9d9d cf2e328 f4e9d9d cf2e328 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 |
"""
Reference
- https://docs.streamlit.io/library/api-reference/layout
- https://github.com/CodingMantras/yolov8-streamlit-detection-tracking/blob/master/app.py
- https://huggingface.co/keremberke/yolov8m-valorant-detection/tree/main
- https://docs.ultralytics.com/usage/python/
"""
import time
import PIL
import streamlit as st
import torch
from ultralyticsplus import YOLO, render_result
from convert import convert_to_braille_unicode, parse_xywh_and_class, braille_to_text
def load_model(model_path):
"""load model from path"""
model = YOLO(model_path)
return model
def load_image(image_path):
"""load image from path"""
image = PIL.Image.open(image_path)
return image
# title
st.title("Braille Pattern Detection")
# sidebar
st.sidebar.header("Detection Config")
conf = float(st.sidebar.slider("Class Confidence", 10, 75, 15)) / 100
iou = float(st.sidebar.slider("IoU Threshold", 10, 75, 15)) / 100
model_path = "snoop2head/yolov8m-braille"
try:
model = load_model(model_path)
model.overrides["conf"] = conf # NMS confidence threshold
model.overrides["iou"] = iou # NMS IoU threshold
model.overrides["agnostic_nms"] = False # NMS class-agnostic
model.overrides["max_det"] = 1000 # maximum number of detections per image
except Exception as ex:
print(ex)
st.write(f"Unable to load model. Check the specified path: {model_path}")
source_img = None
source_img = st.sidebar.file_uploader(
"Choose an image...", type=("jpg", "jpeg", "png", "bmp", "webp")
)
# Single column layout - only show uploaded image
if source_img is None:
default_image_path = "./images/alpha-numeric.jpeg"
image = load_image(default_image_path)
st.image(
default_image_path, caption="Example Input Image", use_column_width=True
)
else:
image = load_image(source_img)
st.image(source_img, caption="Uploaded Image", use_column_width=True)
# Process the image
with st.spinner("Wait for it..."):
start_time = time.time()
try:
with torch.no_grad():
res = model.predict(
image, save=True, save_txt=True, exist_ok=True, conf=conf
)
boxes = res[0].boxes # first image
list_boxes = parse_xywh_and_class(boxes)
except Exception as ex:
st.write("Please upload image with types of JPG, JPEG, PNG ...")
# Convert braille to text and display results
detected_text_lines = []
try:
st.success(f"Done! Inference time: {time.time() - start_time:.2f} seconds")
st.subheader("Detected Text")
for box_line in list_boxes:
str_left_to_right = ""
box_classes = box_line[:, -1]
for each_class in box_classes:
braille_unicode = convert_to_braille_unicode(model.names[int(each_class)])
str_left_to_right += braille_unicode
# Convert braille unicode to actual text
text_line = braille_to_text(str_left_to_right)
detected_text_lines.append(text_line)
st.write(text_line)
# Combine all detected text
full_detected_text = "\n".join(detected_text_lines)
# Add copy to clipboard functionality
if st.button("Copy Text to Clipboard"):
# Use streamlit's built-in clipboard functionality
st.code(full_detected_text, language=None)
st.success("Text displayed above. Use Ctrl+A to select all, then Ctrl+C to copy!")
# Alternative: Create a text area that users can easily copy from
st.subheader("Copy Text Below:")
st.text_area("Detected Text (Select All & Copy)", value=full_detected_text, height=150)
except Exception as ex:
st.write("Please try again with images with types of JPG, JPEG, PNG ...") |