andrei-pokrovsky / rfcn-tensorflow

RFCN implementation in TensorFlow

TensorFlow implementation of RFCN

Paper is available on https://arxiv.org/abs/1605.06409.

Building

The ROI pooling and the MS COCO loader needs to be compiled first. To do so, run make in the root directory of the project. You may need to edit BoxEngine/ROIPooling/Makefile if you need special linker/compiler options.

Testing

You can run trained models with test.py. Model path should be given without file extension (without .data* and .index). An example:

License

The software is under Apache 2.0 license. See http://www.apache.org/licenses/LICENSE-2.0 for further details.

Notes

This code requires TensorFlow 1.0.