jhfjhfj1 / autokeras / 1177

91%

push

travis-ci

[MRG] Object Detection (#396)

* object detection v0.1 added

* add import object_detection in __init__.py under autokeras

* bug fixed

* add support for customized VOC-like dataset

* add load() and predict() in object_detector.py

* update .gitignore

* add download_file support & bug fixed

* update load() and predict(); add pretrained ABC

* update data utils

* update predict

* update predict()

* add test cases

* update get_device and remove code not used

* update load()

* remove redundant code

* test object detection

* pretrain

* bug fixed in load() when no cuda device avaialbe

* comment cudnn & update object detection tests

* bug fix

* bug fixed (#366)

* change

* refactor

* fixed unittests import

355 of 355 new or added lines in 5 files covered. (100.0%)

3682 of 3930 relevant lines covered (93.69%)

0.94 hits per line