comma10k VS lightly

Comma pencil is still a thing that needs lots of help. https://github.com/commaai/comma10k Also, sure read the OpenPilot code, find bugs or improvements and submit useful PRs.

The modern successor is https://github.com/commaai/comma10k

You can see the raw data that Comma is using for both cameras here: https://github.com/commaai/comma10k/tree/master/imgs2

TRY A TWO! https://comma.ai/compare Help with comma pencil! https://github.com/commaai/comma10k

There's also comma10k, which is actually more in the spirit of open source relying on open collaboration. https://github.com/commaai/comma10k

Openpilot is not fully end to end to yet. The longitudinal policy is still done classically, which requires explicit lead car detection. And by default the lateral does explicit lane line detection to keep itself centered with a hand coded planer. I can't comment on exactly how it's done since my understand of machine learning is purely superficial, but the model is also trained to detect lane lines and lead cars using a segnet based on this comma10k dataset, which has hand labeled images. Sky isn't one of these labels though.

This is great for projects where management or tracking of some sort is done from a public Google spreadsheet. Here's some examples where I've put it to use.

Here's a project where I sent in a PR that was merged to display annotation progress:

https://github.com/commaai/comma10k/pull/2767

Here's a wiki entry using it to display the current number of bounty contributors and bounty amount ($8K+!):

https://github.com/commaai/openpilot/wiki/Toyota-Lexus#2021-...

If you're organizing something for your community using a spreadsheet, maybe consider using Cell Shield to put a little badge somewhere visible.

Our goal is to use active learning to use a COCO pre-trained model and fine-tune it on a dataset for autonomous driving. For this transfer task, we are using the Comma10k dataset. From the repository: "It's 10,000 PNGs of real driving captured from the comma fleet. It's MIT license, no academic-only restrictions or anything."

Another year of has passed, and we’ve seen exciting progress in research around self-supervised learning in computer vision. We’re very excited that some of the recent models such as Masked Autoencoders (MAE) or Masked Siamese Networks (MSN) have been added to our OSS framework.

If you’re interested in self-supervised learning and want to try it out yourself you can check out our open-source repository for self-supervised learning.

I assume that you have an image of the invoice. Then using a framework like https://github.com/lightly-ai/lightly with many implemented algorithms is the way to go. And after that step, with model-producing embeddings, you need to compare the embedding of a query with your known database and check if the distance is below some threshold. Of course, pipeline with checking the closest neighbor can be more complicated but I would start with sth really simple.

https://github.com/lightly-ai/lightly implements a lot of self supervised models, and had been available for a while.

modAL indeed has a similar goal of choosing the best subset of data to be labeled. However it has some notable differences:

modAL is built on scikit-learn which is also evident from the suggested workflow. Lightly on the other hand was specifically built for deep learning applications supporting active learning for classification but also object detection and semantic segmentation.

modAL provides uncertainty-based active learning. However, it has been shown that uncertainty-based AL fails at batch-wise AL for vision datasets and CNNs, see https://arxiv.org/abs/1708.00489. Furthermore it only works with an initially trained model and thus labeled dataset. Lightly offers self-supervised learning to learn high dimensional embeddings through its open-source package https://github.com/lightly-ai/lightly. They can be used through our API to choose a diverse subset. Optionally, this sampling can be combined with uncertainty-based AL.

You can easily train, embed, and upload a dataset using the lightly Python package.  First, we need to install the package. We recommend using pip for this. Make sure you're in a Python3.6+ environment. If you're on Windows you should create a conda environment.