资源分类
》 资源 》算法》detectron2-helpers

detectron2-helpers

2020-03-05 | |  89 |   0 |   0

detectron2-helpers

Helpers for detectron2

DummyAlbuMapper allows you to use albumentations during fitting detectron2 for detection tasks.

It requires improvements to work with keypoints and masks.

To use custom dataset_mapper you also need to define new class Trainer with modified build_train_loader. Smth like this:

from detectron2.engine import DefaultTrainerclass Trainer(DefaultTrainer):    @classmethod
    def build_train_loader(cls, cfg):
        mapper = DummyAlbuMapper(cfg, is_train=True)        return build_detection_train_loader(cfg, mapper=mapper)


上一篇：detectron-traffic-signs

下一篇：RetinaNet-and-SSD-in-PyTorch-Detectron

用户评价
全部评价

热门资源

  • DuReader_QANet_BiDAF

    Machine Reading Comprehension on DuReader Usin...

  • ETD_cataloguing_a...

    ETD catalouging project using allennlp

  • honk-models

    Honk models google-speech-dataset.pt: model fo...

  • couleurs-des-van-...

    couleurs-des-van-gogh

  • allennlp_extras

    allennlp_extras Some utilities build on top of...