utility script to evaluate an OD model on a custom dataset - cppe5

scratch_cppe5.py

import os
import json 
import evaluate
import torch
import torchvision
import numpy as np
from tqdm import tqdm
from PIL import Image
from transformers import DetrFeatureExtractor, DetrForObjectDetection, DetrImageProcessor

from datasets import load_dataset

cppe5 = load_dataset("cppe-5")

class CocoDetection(torchvision.datasets.CocoDetection):
    def __init__(self, img_folder, feature_extractor, ann_file):
        super(CocoDetection, self).__init__(img_folder, ann_file)
        self.feature_extractor = feature_extractor

    def __getitem__(self, idx):
        # read in PIL image and target in COCO format
        img, target = super(CocoDetection, self).__getitem__(idx)
        
        # preprocess image and target (converting target to DETR format, resizing + normalization of both image and target)
        image_id = self.ids[idx]
        target = {'image_id': image_id, 'annotations': target}
        encoding = self.feature_extractor(images=img, annotations=target, return_tensors="pt")
        pixel_values = encoding["pixel_values"].squeeze() # remove batch dimension
        target = encoding["labels"][0] # remove batch dimension

        return pixel_values, target

feature_extractor = DetrFeatureExtractor.from_pretrained("MariaK/detr-resnet-50_fine_tuned_cppe5")
im_processor = DetrImageProcessor.from_pretrained("MariaK/detr-resnet-50_fine_tuned_cppe5")
model = DetrForObjectDetection.from_pretrained("MariaK/detr-resnet-50_fine_tuned_cppe5")

def collate_fn(batch):
  pixel_values = [item[0] for item in batch]
  encoding = feature_extractor.pad_and_create_pixel_mask(pixel_values, return_tensors="pt")
  labels = [item[1] for item in batch]
  batch = {}
  batch['pixel_values'] = encoding['pixel_values']
  batch['pixel_mask'] = encoding['pixel_mask']
  batch['labels'] = labels
  return batch

# prepare the test dataset
def val_formatted_anns(image_id, objects):
  annotations = []
  for i in range(0,len(objects["id"])):
    new_ann = {"id": objects["id"][i],
          "category_id": objects["category"][i],
          "iscrowd": 0,
          "image_id": image_id,
          "area": objects["area"][i],
          "bbox": objects["bbox"][i]}
    annotations.append(new_ann)
    
  return annotations 

def save_cppe5_annotation_file_images(cppe5):
  output_json = {}
  path_output_cppe5 = f"{os.getcwd()}/cppe5/"

  if not os.path.exists(path_output_cppe5):
    os.makedirs(path_output_cppe5)

  path_anno = os.path.join(path_output_cppe5, "cppe5_ann.json")
  # We map them to their category following the COCO format
  # 0: coverall
  # 1: face_shield
  # 2: gloves
  # 3: goggles
  # 4: mask
  categories_json = [
    {
      "supercategory": "none",
      "id": 0,
      "name": "coverall"
    },
    {
      "supercategory": "none",
      "id": 1,
      "name": "face_shield",
    },
    {
      "supercategory": "none",
      "id": 2,
      "name": "gloves",
    },
    {
      "supercategory": "none",
      "id": 3,
      "name": "goggles",
    },
    {
      "supercategory": "none",
      "id": 4,
      "name": "mask",
    }
  ]
  output_json["images"] = []
  output_json["annotations"] = []
  for batch in cppe5:
      ann = val_formatted_anns(batch["image_id"], batch["objects"])
      output_json["images"].append(
        {
          "id": batch["image_id"],
          "width": batch["image"].width,
          "height": batch["image"].height,
          "file_name": f"{batch['image_id']}.png"
        }
      )
      output_json["annotations"].extend(ann)
  output_json["categories"] = categories_json

  with open(path_anno, "w") as file:
      json.dump(output_json, file, ensure_ascii=False, indent=4)


  for im, img_id in zip(cppe5["image"], cppe5["image_id"]):
      path_img = os.path.join(path_output_cppe5, f"{img_id}.png")
      im.save(path_img)
  
  return path_output_cppe5, path_anno

path_output_cppe5, path_anno = save_cppe5_annotation_file_images(cppe5["test"])

dummy = CocoDetection(path_output_cppe5, feature_extractor, path_anno)
module = evaluate.load("ybelkada/cocoevaluate", coco=dummy.coco)
val_dataloader = torch.utils.data.DataLoader(dummy, batch_size=8, shuffle=False, num_workers=4, collate_fn=collate_fn)

with torch.no_grad():
  for idx, batch in enumerate(tqdm(val_dataloader)):
      # set to device 
      pixel_values = batch["pixel_values"]
      pixel_mask = batch["pixel_mask"]

      labels = [{k: v for k, v in t.items()} for t in batch["labels"]] # these are in DETR format, resized + normalized

      # forward pass
      outputs = model(pixel_values=pixel_values, pixel_mask=pixel_mask)

      orig_target_sizes = torch.stack([target["orig_size"] for target in labels], dim=0)
      results = im_processor.post_process(outputs, orig_target_sizes) # convert outputs of model to COCO api

      module.add(prediction=results, reference=labels)
      del batch

results = module.compute()
print(results)