main file.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "83bb3cb1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model loaded successfully from yolo/weights/yolov3-df2_15000.weights.\n",
      "img path: tests/11.jpg\n",
      "\t+ Label: short sleeve dress, Conf: 0.54085\n",
      "length:250.0\n",
      "\t+ Label: short sleeve dress, Conf: 0.55985\n",
      "length:599.0\n",
      "\t+ Label: shorts, Conf: 0.60323\n",
      "length:168.0\n",
      "\t+ Label: long sleeve outwear, Conf: 0.72779\n",
      "length:363.0\n",
      "\t+ Label: long sleeve outwear, Conf: 0.80595\n",
      "length:355.0\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "import os\n",
    "import cv2\n",
    "from yolo.utils.utils import *\n",
    "from predictors.YOLOv3 import YOLOv3Predictor\n",
    "import glob\n",
    "from tqdm import tqdm\n",
    "import sys\n",
    "\n",
    "\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "torch.cuda.empty_cache()\n",
    "\n",
    "\n",
    "#YOLO PARAMS\n",
    "yolo_df2_params = { \"model_def\" : \"yolo/df2cfg/yolov3-df2.cfg\",\n",
    "\"weights_path\" : \"yolo/weights/yolov3-df2_15000.weights\",\n",
    "\"class_path\":\"yolo/df2cfg/df2.names\",\n",
    "\"conf_thres\" : 0.5,\n",
    "\"nms_thres\" :0.4,\n",
    "\"img_size\" : 416,\n",
    "\"device\" : device}\n",
    "\n",
    "\n",
    "#DATASET\n",
    "dataset = 'deepfashion2' #deepfashion2\n",
    "yolo_params = yolo_df2_params\n",
    "\n",
    "#Classes\n",
    "classes = load_classes(yolo_params[\"class_path\"])\n",
    "\n",
    "#Colors\n",
    "cmap = plt.get_cmap(\"rainbow\")\n",
    "colors = np.array([cmap(i) for i in np.linspace(0, 1, 13)])\n",
    "\n",
    "model = 'yolo'\n",
    "detectron = YOLOv3Predictor(params=yolo_params)\n",
    "\n",
    "while(True):\n",
    "    path = input('img path: ')\n",
    "    if not os.path.exists(path):\n",
    "        print('Img does not exists..')\n",
    "        continue\n",
    "    img = cv2.imread(path)\n",
    "    detections = detectron.get_detections(img)\n",
    "\n",
    "    \n",
    "    if len(detections) != 0 :\n",
    "        detections.sort(reverse=False ,key = lambda x:x[4])\n",
    "        for x1, y1, x2, y2, cls_conf, cls_pred in detections:\n",
    "\n",
    "                print(\"\\t+ Label: %s, Conf: %.5f\" % (classes[int(cls_pred)], cls_conf))           \n",
    "                color = colors[int(cls_pred)]\n",
    "                \n",
    "                color = tuple(c*255 for c in color)\n",
    "                color = (.7*color[2],.7*color[1],.7*color[0])       \n",
    "                    \n",
    "                font = cv2.FONT_HERSHEY_SIMPLEX   \n",
    "            \n",
    "            \n",
    "                x1, y1, x2, y2 = int(x1), int(y1), int(x2), int(y2)\n",
    "                text =  \"%s conf: %.1f\" % (classes[int(cls_pred)] ,cls_conf)\n",
    "                \n",
    "                cv2.circle(img,(x1,y1),8,(0,0,255),-1)\n",
    "                cv2.circle(img,(x1,y2),8,(0,0,255),-1)\n",
    "                cv2.line(img,(x1,y1),(x1,y2),color,5)\n",
    "                distance = ((((x1 - x1 )**2) + ((y2-y1)**2) )**0.5) \n",
    "                print(\"length:{:}\".format(distance))\n",
    "                \n",
    "                y1 = 0 if y1<0 else y1\n",
    "                y1_rect = y1-25\n",
    "                y1_text = y1-5\n",
    "\n",
    "                if y1_rect<0:\n",
    "                    y1_rect = y1+27\n",
    "                    y1_text = y1+20\n",
    "                cv2.putText(img,text,(x1,y1_text), font, 0.5,(255,0,0),1,cv2.LINE_AA)\n",
    "                cv2.putText(img,\"length: {:}\".format(str(distance)),(x1,y2+20), font, 0.5,(0,0,255),1,cv2.LINE_AA)\n",
    "                \n",
    "            \n",
    "    cv2.imshow('length Detection',cv2.resize(img,(1500,600) ))  \n",
    "    img_id = path.split('/')[-1].split('.')[0]\n",
    "    cv2.imwrite('output/ouput-test_{}_{}_{}.jpg'.format(img_id,model,dataset),img)\n",
    "    \n",
    "    cv2.waitKey(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ed6f70ab",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}