Edgar Riba

initial commit for cli

2

#### Changes Fixes # (issue) #### Type of change - [ ] 📚 Documentation Update - [ ] 🧪 Tests Cases - [ ] 🐞 Bug fix (non-breaking change which...

implement `kornia_imgproc::color::rgb_from_grayscale` -- stack the color pixel 3x ```rust img_rgb8u .as_slice_mut() .chunks_exact_mut(3) .zip(img_mono8u.as_slice()) .for_each(|(chunk, &pixel)| { chunk[0] = pixel; chunk[1] = pixel; chunk[2] = pixel; }); ```

Currently we rely on `tokio` to handle the async mechanism to push/pull data from gstreamer. However, when integrating with other frameworks like [`nodo`](https://github.com/Danvil/nodo) which do not implement by default the...

Implement VideoWriter / VideoReader

1

implement an initial mp4 reader/writer in kornia-io Evaluate wheter to use the crates below - gstreamer: https://github.com/kornia/kornia-rs/blob/video/src/io/video.rs [this solution for now as can support hardware acceleration] - ~~https://github.com/alfg/mp4-rust~~ also requested...

Implement a crop function in the kornia imgproc. - Possibly this requires some sort of slice function in kornia_core::Tensor. - Evaluate whether this can be implemented as a view

Add support for `nvcodec` in the `StreamCapture`

2

Implement hardware acceleration for the `StreamCapture` via `nvcodec` *https://gstreamer.freedesktop.org/documentation/nvcodec/index.html?gi-language=c This means extending the `Gstreamer` pipeline for capture as gstreamer has support already for it.

Follow the implementation below  *https://www.linkedin.com/posts/ashvardanian_100x-10000x-numerical-error-reduction-activity-7242002088803663872-kVsx?utm_source=share&utm_medium=member_desktop

Add sccache support in the CI

3

Use `sccache` in the CI in order to bootstrap compilation time - https://crates.io/crates/sccache

Add support for writing images with async that can be useful for i/o bound tasks in robotics and avoid people creating their threadpool. requested by @haixuanTao for [dora-rs](https://github.com/dora-rs/dora) Current impl:...

- creates `kornia-dnn` based on the [`ort`](https://ort.pyke.io/) crate - define `RTDETRDetectorBuilder` and `RTDETRDetector` create the ort session - create a basic `Detection` struct message - add example `rtdetr` to use...