Guan Ting, Lin

Low evaluation result with official model(ZF)

3

Hi All, I evaluate the official ZF-net model which is downloaded with fetch_faster_rcnn_models.sh script. My faster rcnn is download from "https://github.com/rbgirshick/py-faster-rcnn". Testing command are: ./tools/test_net.py --net data/faster_rcnn_models/ZF_faster_rcnn_final.caffemodel --def models/pascal_voc/ZF/faster_rcnn_alt_opt/faster_rcnn_test.pt --cfg...

Weird result with your upload model and word map

2

Hi, Thank you for the detailed tutorial, it helps me a lot. When I use your model to evaluate the performance on COCO data set. It return the bleu score...

Low training accuracy on cityscape data set

2

Hi, I am training the caffe version of ENet. When training on the cityscape fine annotated dataset, my training accuracy seems not to be converged. Do you have any suggestion...