| import json |
| import random |
| from tqdm import tqdm |
| DEFAULT_IMAGE_TOKEN = '<image>' |
|
|
| SHORT_QUESTION = [ |
| "Can you segment the {class_name} in this image?", |
| "Please segment the {class_name} in this image.", |
| "What is {class_name} in this image? Please respond with segmentation mask.", |
| "What is {class_name} in this image? Please output segmentation mask.", |
| "Could you identify and segment the {class_name} in this image?", |
| "Would you be able to segment the {class_name} in this image?", |
| "Can you provide a segmentation mask for the {class_name} in this image?", |
| "Please provide a segmentation mask for the {class_name} in this image.", |
| "Could you please segment the {class_name} in this image for me?", |
| "What {class_name} is present in this image? Kindly respond with a segmentation mask.", |
| "Which part of this image contains {class_name}? Please output with segmentation mask.", |
| "Is there a {class_name} in this image? If so, please provide the segmentation mask.", |
| "Can you segment out the {class_name} visible in this image?", |
| "Would you identify and provide a segmentation mask for the {class_name} in this image?", |
| ] |
| |
| ANSWER_LIST = [ |
| "It is [SEG].", |
| "Sure, [SEG].", |
| "Sure, it is [SEG].", |
| "Sure, the segmentation result is [SEG].", |
| "[SEG].", |
| ] |
|
|
| NO_MASK_LIST = [ |
| "Sorry, there is no {class_name} in the image.", |
| "There is no {class_name} in the image.", |
| "It seems that there is no {class_name} in the image.", |
| "It appears that the image doesn't contain any {class_name}.", |
| "Unfortunately, I'm unable to find any {class_name} in the image.", |
| "I regret to inform you that there is no {class_name} detected in the image.", |
| "The image seems to lack any presence of {class_name}.", |
| "It looks like there are no {class_name} visible in the image.", |
| "I'm sorry, but there doesn't appear to be any {class_name} in the image.", |
| "It seems like the image does not include any {class_name}.", |
| "I'm afraid there is no sign of {class_name} in the image.", |
| "Regrettably, the image does not feature any {class_name}.", |
| "There doesn't seem to be any {class_name} found in the image.", |
| ] |
| |
| data_list = [ |
| 'refcoco.json', |
| 'refcocog.json', |
| 'refcoco+.json', |
| |
| ] |
|
|
| img_dic = {} |
| final_data = [] |
| json_data = json.load(open('grefcoco.json')) |
| for data in json_data: |
| img_dic[data['image']] = data |
|
|
| for d_ in tqdm(data_list): |
| json_data = json.load(open(d_)) |
| for data in tqdm(json_data): |
| if data['image'] in img_dic: |
| for i in range(len(data['cat'])): |
| if data['cat'][i] in img_dic[data['image']]['cat']: |
| continue |
| img_dic[data['image']]['cat'].append(data['cat'][i]) |
| img_dic[data['image']]['masks'].append([data['masks'][i]]) |
| |
| idx = 0 |
| for k in tqdm(img_dic): |
| data = img_dic[k] |
| if len(data['masks'])!=len(data['cat']): |
| print('err') |
| indices = list(range(len(data['masks']))) |
|
|
| |
| random.shuffle(indices) |
|
|
| data['masks'] = [data['masks'][i] for i in indices] |
| data['cat'] = [data['cat'][i] for i in indices] |
| |
| s = 0 |
| while s<len(data['masks']): |
| num = random.randint(2, 5) |
| dic = {} |
| dic['id'] = f'ref_{idx}' |
| idx+=1 |
| dic['image'] = data['image'] |
| dic['height'] = data['height'] |
| dic['width'] = data['width'] |
| dic['conversations'] = [] |
| dic['masks'] = [] |
| for i in range(num): |
| if i+s>=len(data['masks']): |
| break |
| dic['conversations'].append({'from': 'human', 'value': random.choice(SHORT_QUESTION).format(class_name=data['cat'][i+s].lower())}) |
| if data['masks'][i+s] is None: |
| dic['conversations'].append({'from': 'gpt', 'value': random.choice(NO_MASK_LIST).format(class_name=data['cat'][i+s].lower())}) |
| else: |
| dic['conversations'].append({'from': 'gpt', 'value': random.choice(ANSWER_LIST).replace('[SEG]', '[SEG]'*len(data['masks'][i+s]))}) |
| for ann in data['masks'][i+s]: |
| dic['masks'].append(ann) |
| dic['conversations'][0]['value'] = '<image>\n' + dic['conversations'][0]['value'] |
| s+=num |
| final_data.append(dic) |
| print(len(final_data)) |
|
|
| with open('ref_all.json', 'w') as f: |
| f.write(json.dumps(final_data, indent=4)) |
|
|
