EveryDream/scripts/auto_caption.py

import argparse
import glob
import os
from PIL import Image
import sys
from torchvision import transforms
from torchvision.transforms.functional import InterpolationMode
import torch
import aiohttp
import asyncio
import subprocess
import numpy as np
import io
import aiofiles

SIZE = 384
BLIP_MODEL_URL = 'https://storage.googleapis.com/sfr-vision-language-research/BLIP/models/model_base_caption_capfilt_large.pth'

def get_parser(**parser_kwargs):
    parser = argparse.ArgumentParser(**parser_kwargs)
    parser.add_argument(
        "--img_dir",
        type=str,
        nargs="?",
        const=True,
        default="input",
        help="directory with images to be captioned",
    ),
    parser.add_argument(
        "--out_dir",
        type=str,
        nargs="?",
        const=True,
        default="output",
        help="directory to put captioned images",
    ),
    parser.add_argument(
        "--format",
        type=str,
        nargs="?",
        const=True,
        default="filename",
        help="'filename', 'mrwho', 'txt', or 'caption'",
    ),
    parser.add_argument(
        "--nucleus",
        type=bool,
        nargs="?",
        const=True,
        default=False,
        help="use nucleus sampling instead of beam",
    ),
    parser.add_argument(
        "--q_factor",
        type=float,
        nargs="?",
        const=True,
        default=1.0,
        help="adjusts the likelihood of a word being repeated",
    ),
    parser.add_argument(
        "--min_length",
        type=int,
        nargs="?",
        const=True,
        default=22,
        help="adjusts the likelihood of a word being repeated",
    ),
    parser.add_argument(
        "--torch_device",
        type=str,
        nargs="?",
        const=False,
        default="cuda",
        help="specify a different torch device, e.g. 'cpu'",
    ),

    return parser

def load_image(raw_image, device):
    transform = transforms.Compose([
        #transforms.CenterCrop(SIZE),
        transforms.Resize((SIZE, SIZE), interpolation=InterpolationMode.BICUBIC),
        transforms.ToTensor(),
        transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))
    ])
    image = transform(raw_image).unsqueeze(0).to(device)
    return image

def get_out_file_name(out_dir, base_name, ext):
    return os.path.join(out_dir, f"{base_name}{ext}")

async def main(opt):
    print("starting")
    import models.blip

    sample = False
    if opt.nucleus:
        sample = True

    input_dir = opt.img_dir
    print("input_dir: ", input_dir)

    config_path = "scripts/BLIP/configs/med_config.json"

    cache_folder = ".cache"
    model_cache_path = ".cache/model_base_caption_capfilt_large.pth"

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

    if not os.path.exists(opt.out_dir):
        os.makedirs(opt.out_dir)

    if not os.path.exists(model_cache_path):
        print(f"Downloading model to {model_cache_path}... please wait")

        async with aiohttp.ClientSession() as session:
            async with session.get(BLIP_MODEL_URL) as res:
                with open(model_cache_path, 'wb') as f:
                    async for chunk in res.content.iter_chunked(1024):
                        f.write(chunk)
        print(f"Model cached to: {model_cache_path}")
    else:
        print(f"Model already cached to: {model_cache_path}")

    blip_decoder = models.blip.blip_decoder(pretrained=model_cache_path, image_size=SIZE, vit='base', med_config=config_path)
    blip_decoder.eval()

    print(f"loading model to {opt.torch_device}")

    blip_decoder = blip_decoder.to(torch.device(opt.torch_device))

    ext = ('.jpg', '.jpeg', '.png', '.webp', '.tif', '.tga', '.tiff', '.bmp', '.gif')

    i = 0

    for idx, img_file_name in enumerate(glob.iglob(os.path.join(opt.img_dir, "*.*"))):
        if img_file_name.endswith(ext):
            caption = None
            file_ext = os.path.splitext(img_file_name)[1]
            if (file_ext in ext):
                async with aiofiles.open(img_file_name, "rb") as input_file:
                    print("working image: ", img_file_name)

                    image_bin = await input_file.read()
                    image = Image.open(io.BytesIO(image_bin))

                    if not image.mode == "RGB":
                        image = image.convert("RGB")

                    image = load_image(image, device=torch.device(opt.torch_device))

                    if opt.nucleus:
                        captions = blip_decoder.generate(image, sample=True, top_p=opt.q_factor)
                    else:
                        captions = blip_decoder.generate(image, sample=sample, num_beams=16, min_length=opt.min_length, \
                            max_length=48, repetition_penalty=opt.q_factor)

                    caption = captions[0]

                    if opt.format in ["mrwho","joepenna"]:
                        prefix = f"{i:05}@"
                        i += 1
                        caption = prefix+caption
                    elif opt.format == "filename":
                        postfix = f"_{i}"
                        i += 1
                        caption = caption+postfix

                    if opt.format in ["txt","text","caption"]:
                        out_base_name = os.path.splitext(os.path.basename(img_file_name))[0]

                    if opt.format in ["txt","text"]:
                        out_file = get_out_file_name(opt.out_dir, out_base_name, ".txt")

                    if opt.format in ["caption"]:
                        out_file = get_out_file_name(opt.out_dir, out_base_name, ".caption")

                    if opt.format in ["txt","text","caption"]:
                        print("writing caption to: ", out_file)
                        async with aiofiles.open(out_file, "w") as out_file:
                            await out_file.write(caption)

                    if opt.format in ["filename", "mrwho", "joepenna"]:
                        caption = caption.replace("/", "").replace("\\", "")  # must clean slashes using filename
                        out_file = get_out_file_name(opt.out_dir, caption, file_ext)
                        async with aiofiles.open(out_file, "wb") as out_file:
                            await out_file.write(image_bin)
                    elif opt.format == "json":
                        raise NotImplementedError
                    elif opt.format == "parquet":
                        raise NotImplementedError

def isWindows():
    return sys.platform.startswith("win")

if __name__ == "__main__":
    parser = get_parser()
    opt = parser.parse_args()

    if opt.format not in ["filename", "mrwho", "joepenna", "txt", "text", "caption"]:
        raise ValueError("format must be 'filename', 'mrwho', 'txt', or 'caption'")

    if (isWindows()):
        print("Windows detected, using asyncio.WindowsSelectorEventLoopPolicy")
        asyncio.set_event_loop_policy(asyncio.WindowsSelectorEventLoopPolicy())
    else:
        print("Unix detected, using default asyncio event loop policy")

    if not os.path.exists("scripts/BLIP"):
        print("BLIP not found, cloning BLIP repo")
        subprocess.run(["git", "clone", "https://github.com/salesforce/BLIP", "scripts/BLIP"])
    blip_path = "scripts/BLIP"
    sys.path.append(blip_path)

    asyncio.run(main(opt))