tools/object_detection_benchmark.py

# Copyright (c) Facebook, Inc. and its affiliates.

# This source code is licensed under the MIT license found in the
# LICENSE file in the root directory of this source tree.

"""
Full credits: https://github.com/facebookresearch/moco/blob/main/detection/train_net.py  # NOQA
"""


import os

from detectron2.checkpoint import DetectionCheckpointer
from detectron2.config import get_cfg
from detectron2.engine import (
    default_argument_parser,
    default_setup,
    DefaultTrainer,
    launch as d2_launch,
)
from detectron2.evaluation import COCOEvaluator, PascalVOCDetectionEvaluator
from detectron2.layers import get_norm
from detectron2.modeling.roi_heads import Res5ROIHeads, ROI_HEADS_REGISTRY


@ROI_HEADS_REGISTRY.register()
class Res5ROIHeadsExtraNorm(Res5ROIHeads):
    """
    As described in the MOCO paper, there is an extra BN layer
    following the res5 stage.
    """

    def __init__(self, cfg, input_shape):
        super().__init__(cfg, input_shape)
        norm = cfg.MODEL.RESNETS.NORM
        norm = get_norm(norm, self.res5[-1].out_channels)
        self.res5.add_module("norm", norm)


class Trainer(DefaultTrainer):
    @classmethod
    def build_evaluator(cls, cfg, dataset_name, output_folder=None):
        if output_folder is None:
            output_folder = os.path.join(cfg.OUTPUT_DIR, "inference")
        if "coco" in dataset_name:
            return COCOEvaluator(dataset_name, cfg, True, output_folder)
        else:
            assert "voc" in dataset_name, "Dataset must be coco or voc"
            return PascalVOCDetectionEvaluator(dataset_name)


def main(args):
    # setup the config file
    cfg = get_cfg()
    cfg.merge_from_file(args.config_file)
    cfg.merge_from_list(args.opts)
    cfg.freeze()
    default_setup(cfg, args)

    if args.eval_only:
        model = Trainer.build_model(cfg)
        DetectionCheckpointer(model, save_dir=cfg.OUTPUT_DIR).resume_or_load(
            cfg.MODEL.WEIGHTS, resume=args.resume
        )
        res = Trainer.test(cfg, model)
        return res

    trainer = Trainer(cfg)
    trainer.resume_or_load(resume=args.resume)
    return trainer.train()


if __name__ == "__main__":
    args = default_argument_parser().parse_args()
    print("Arguments:", args)
    d2_launch(
        main,
        args.num_gpus,
        num_machines=args.num_machines,
        machine_rank=args.machine_rank,
        dist_url=args.dist_url,
        args=(args,),
    )