Segmentation-models-pytorch

Updates
- New architecture: MANet (310)
- New encoders from `timm`: mobilenetv3 (355) and gernet (344)
- New loss functions in `smp.losses` module (`smp.utils.losses` would be deprecated in future versions)
- New pretrained weight initialization for first convolution if `in_channels > 3`
- Updated timm version (0.4.12)
- Bug fixes and docs improvement

Thanks to azkalot1 JulienMaille originlake Kupchanski loopdigga96 zurk nmerty ludics Vozf markson14 and others!

Updates
- New architecture Unet++ (279)
- New encoders RegNet, ResNest, SK-Net, Res2Net (286)
- Updated timm version (0.3.2)
- Improved docstrings and typehints for models
- Project documentation on https://smp.readthedocs.io

Thanks to azkalot1 for the new encoders and architecture!

Fixes
- Fix `pytorch-efficientnet` package version in requirements.txt to strict 0.6.3 (260)

Updates
- New decoders DeepLabV3, DeepLabV3+, PAN
- New backbones (encoders) `timm-efficientnet*`
- New pretrained weights (ssl, wsl) for resnets
- New pretrained weights (advprop) for efficientnets

And some small fixes.

Thanks IlyaDobrynin gavrin-s lizmisha suitre77 thisisiron phamquiluan and all other contributers!

Updates

1) New backbones (mobilenet, efficientnet, inception)
2) `depth` and `in_channels` options for all models
3) Auxiliary classification output

Note!
Model architectures have been changed, use previous versions for weights compatibility!

Updates
- Conv2D Initialization
- kaiming_normal -> kaiming_uniform;
- fan_out -> fan_in;
- bias -> 0
- package dependencies