GithubHelp home page GithubHelp logo

valeoai / bevcontrast Goto Github PK

View Code? Open in Web Editor NEW
60.0 60.0 5.0 370 KB

BEVContrast: Self-Supervision in BEV Space for Automotive Lidar Point Clouds - Official PyTorch implementation

License: Other

Dockerfile 0.67% Python 99.33%

bevcontrast's People

Contributors

csautier avatar

Stargazers

avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar

Watchers

avatar avatar avatar avatar avatar avatar

Forkers

eaphan 3d-pretraining point-cloud-upup multi-molidaty-3d-perception

bevcontrast's Issues

Question about the BEV projection of the extracted feature

Thank you for this excellent work.

One thing make me confused.

BEVContrast/bevlab/models.py

Line 54 in 54497c9

input_bev = self.collapse_to_bev(input_fmap, batch['coordinates_in'], batch["pc_min_in"].cuda(), self.range, self.voxel_size, self.bev_stride)

  1. Why project features onto a BEV image and then perform rotation and translation?

  2. Would it be more convenient to rotate and translate the point cloud and then project the feature into the BEV image? It also does not need to resample the feature by F.affine_grid and F.grad_sample later.

    BEVContrast/bevlab/models.py

    Line 98 in 54497c9

    grid = F.affine_grid(P.to(device, non_blocking=True), output_bev.shape, align_corners=False)

    BEVContrast/bevlab/models.py

    Line 100 in 54497c9

    pred_bev = F.grid_sample(input_bev[0::self.input_frames], grid, mode='bilinear', padding_mode='zeros', align_corners=False)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.