RT by @tak_yamm: Yes, you can use it in this manner. @cebraAI has a DeepLabCut loading function already, see docs! https://cebra.ai/docs/usage.html - We demo this use in Ye et al 2023a: https://arxiv.org/abs/2203.07436 - And in Ye et al 2023b AmadeusGPT: https://arxiv.org/abs/2307.04858

R to @TrackingActions: So exciting! Thank you!

RT by @rgonzalezlaiz: 🎉🔥 CEBRA has hit 10K downloads in < 6 months, and we have a new pypi release 0.3.0 to celebrate! 🎉 Full notes: https://github.com/AdaptiveMotorControlLab/CEBRA/releases thanks to @stes_io @NasFilippova @rgonzalezlaiz @celia_bqt @TrackingActions & the new external contributors too 🙏🏼🔥🦓

R to @TrackingActions: Also in Switzerland! 😆

R to @TrackingActions: 🤔🤔

R to @sofroniewn: I’m just not a fan of autoencoders when contrastive learning does a better job at finding true latents; reconstructing complex inputs is still hard. So, in short yes, we start by using identifiable models with @cebraAI and we will have more work up soon on explainability 😃

R to @TrackingActions: Link!! https://www.linkedin.com/posts/mackenziemathis_the-mathis-lab-of-adaptive-motor-control-activity-7115645199027957760-j8Ej?utm_source=share&utm_medium=member_ios

R to @ChuckleScience: Students learnt to use open source tools like @Prusa3D, @arduino, @bonsai_rx, @ThePSF, @DeepLabCut, #keypointMoSeq, @cebraAI, and important concepts to use these in the study of animal behavior.

👀 we have a new pypi release coming out in the next days!! Stay tuned for some nice updates! 🦓🎉

Hello @cebraAI, you have been featured on our portal @AIBleak https://aibleak.com/product/cebra Share with your friends and get reviews to get into the top list.