interpret.py

# --------------------------------------------------------------------------------
# Analyze super-resolution models using LAM.
# Official GitHub: https://github.com/X-Lowlevel-Vision/LAM_Demo
#
# Modified by Jinpeng Shi (https://github.com/jinpeng-s)
# --------------------------------------------------------------------------------
import logging
import os.path
from os import path as osp

from basicsr.models import build_model
from basicsr.utils import get_env_info
from basicsr.utils import get_root_logger
from basicsr.utils import get_time_str
from basicsr.utils.options import dict2str

import archs  # noqa
import data  # noqa
import models  # noqa
from utils import get_model_interpretation
from utils import make_exp_dirs
from utils import parse_options


def interpret_pipeline(root_path):  # noqa
    # parse options, set distributed setting, set random seed
    opt, _ = parse_options(root_path, is_train=False)

    # mkdir and initialize loggers
    make_exp_dirs(opt)
    log_file = osp.join(opt['path']['log'],
                        f"interpret_{opt['name']}_{get_time_str()}.log")
    logger = get_root_logger(logger_name='basicsr',
                             log_level=logging.INFO, log_file=log_file)
    logger.info(get_env_info())
    logger.info(dict2str(opt))

    # create model
    model = build_model(opt)

    logger.info(f'Interpreting {model.net_g.__class__.__name__}...')

    for _, img_opt in sorted(opt['interpret_imgs'].items()):
        img, di = get_model_interpretation(model.net_g, img_opt['img_path'], img_opt['w'], img_opt['h'],
                                           use_cuda=True if opt['num_gpu'] > 0 else False)
        os.makedirs(osp.join(opt['path']['visualization'],
                    'interpretation'), exist_ok=True)
        img.save(osp.join(opt['path']['visualization'],
                 'interpretation', os.path.basename(img_opt['img_path'])))
        logger.info(
            f"DI of {os.path.basename(img_opt['img_path'])}: {round(di, 3)}")
        logger.info("The LAM result are saved to "
                    f"{osp.join(opt['path']['visualization'], 'interpretation')}.")


if __name__ == '__main__':
    root_path = osp.abspath(osp.join(__file__, osp.pardir))
    interpret_pipeline(root_path)