simpledet
simpledet copied to clipboard
Future Roadmap Suggestions
- Rewrite proposal generation from CUDA to Python Thanks to https://github.com/apache/incubator-mxnet/pull/14363, MXNet CustomOp now works parallelly. This enables us to migrate more CUDA codes to Python to ease the modification of core operators.
- [x] Fix RoiAlignv2 compilation failure introduced by https://github.com/apache/incubator-mxnet/pull/14270 in rogerchen/cuda9.0-cudnn7-py3:simpledet
- [x] TVM deployment
- [ ] multi-node sync bn
- [ ] horovod distributed training
- [ ] speedup loader with https://github.com/apache/incubator-mxnet/pull/14733
Rotation Invariant Object Detection to train on DOTA dataset.
- [ ] Support NDArray and Symbol dual modes for user-friendly debug.
- [ ] refactor code with https://github.com/apache/incubator-mxnet/pull/14638
- [ ] refactor code with https://github.com/apache/incubator-mxnet/pull/13985
- [ ] file corruption check for #182
- [ ] wrap the contrib RoIAlign for our use Our one is written in quite an early day and is slightly different from the official Detectron implementation. Also, ours lacks FP16 support, sampling ratio, and PS variant.
pinning memory for loader
- [ ] model parallelism with pipeline