I am building a perception pipeline from RGB-D to do pick-and-place with a depth camera.

In the previous text I described how I get a stable yaw using PCA on the table plane and a circular mean to deal with 180 degree ambiguity.

This time, I want to write about two new insights.

https://medium.com/p/df8de460a8c7?postPublishedType=initial

#robotics #computervision #3D #geometry #maths #rgbd #depthcamera

Yaw, symmetry and a moving camera: two lessons from a tabletop RGB-D grasping pipeline.

In the previous text I described how I get a stable yaw using PCA on the table plane and a circular mean to deal with 180 degree ambiguity.

Medium
In case you missed it: We had a great talk about DroneBlocks, an affordable, comprehensive, drone kit for STEM. Some great audience questions, and a pretty deep look at how they keep costs down and quality up. https://youtube.com/live/P5eingrUQFk #OpenCV #ComputerVision #AI #Robotics #Drones
Revolutionary Drone STEM Kit from DroneBlocks

YouTube
Episode 162 – The robot doctor will see you now - Robot Talk

In this special live recording at the Great Exhibition Road Festival, Claire chatted to George Mylonas, Antonia Tzemanaki and Tom Vercauteren about robotics and AI in medicine and healthcare.

Robot Talk - The podcast exploring the exciting world of robotics, artificial intelligence and autonomous machines.

Computer vision projects feel impressive, emotional, and often score higher.

#computervision #machinelearning #ai

📲 Computer vision is a privacy-degrading corporate activity. Maintaining Freedom of Thought, privacy, dignity and self-determination requires deliberate rejection of surveillance tech that multiplies "computer vision" capabilities in our day-to-day lives.

#privacy #privacymatters #panopticon #surveillance #surveillancecapitalism #freedomofthought #computervision

🖼️ Call a spade a spade: How to deal with objects that are incorrectly/ambiguously labelled in #ML?

New #RKI research from #CVPR2026 outperforms SotA methods by filtering erroneous labels with a “drainage” class.

🔗 https://cvpr.thecvf.com/virtual/2026/poster/39457

#ZKIPH #AI #ComputerVision #AIResearch #ImageClassification

CVPR Poster Drainage: A Unifying Framework for Addressing Class Uncertainty

Since it remains unclear whether my current #PostDoc contract will be prolonged in January, I am starting to look for a new #AcademicJob!

My expertise are #ecology, #ethology, #AnimalBehavior, with a focus on #olfaction in #mammals.
I had a stint in #NeuroEthology, but that wasn't quite my cup of tea, and I am currently working in #AnimalWelfare for husbandry animals (think cows, pigs etc).
I have done field work (wild rodents), lab work (very controlled behavioral assays), and pure data analysis.
I do everything in #rstats, so my standard analysis tools are GLM/GLMM, but I'm currently learning how to properly use Generalized Additive Models, as they seem like the natural next step.
I've also dabbled in #python in order to work with #DeepLabCut and #moseq, so the last few years I spent mostly with #ComputerVision, #DeepLearning, and extracting data from videos.
As that requires heavy GPU calculations, I've done most on my work on institutional and/or national #HPC, and then crunched the masses of resulting data in #rstats

I'm looking for a job in the #EU, preferably anywhere north of Germany, but let's be honest, I'll go wherever there is a job.

My website is a bit outdated, but https://tsievert.com/

Please boost for visibility!

#AnimalBehaviour #FediJobs #FediHire #GetFediHired #LookingForJob

Thorbjörn Sievert, PhD - Personal website

Researcher focused on behavioral ecology, chemical ecology, and predator-prey interactions

NeoEyes NE503 – A $1199 Edge AI Camera based on Hailo-15H 20 TOPS SoC

Camthink NeoEyes NE503 is a 4K Edge AI camera platform powered by the Hailo-15H quad-core Cortex-A53 SoC with a 20 TOPS AI accelerator, and paired with 8GB of LPDDR4 memory and a 64GB eMMC flash. The camera also features a Sony IMX678 4K sensor and an 8–32mm F1.6 AF auto-zoom lens for high-quality monitoring, low-light imaging, and long-range recognition. The camera directly handles high-resolution video input, local multi-model inference, containerized applications, and event-driven workflows on the device, without relying on the cloud or a local server. NeoEyes NE503 specifications: SoC - Hailo-15H CPU - Quad-core Arm Cortex-A53 @  1.3 GHz VPU - H.265/H.264 encoding ISP - Supports up to 12 MP resolution, 600 Mpixel/s pixel rate, HDR, and noise reduction AI performance - Up to 20 TOPS; < 50 ms target inference latency System Memory - 8GB LPDDR4 @ 4266 MT/s; 8.5 GB/s single-channel bandwidth Storage 64 GB eMMC flash 8

CNX Software - Embedded Systems News

🚨 NEWS: Machine Learning con Python — Modelli Pronti per il Business, Costi Controllati e Deployment Reale

Ecco i punti chiave in breve:
💡 Quali sono i fondamenti del Machine Learning con Python per un progetto reale?Se hai mai provato a prevedere le vendite del prossimo trimestre o a classificare ticket di supporto senza far impazzire i...

🚀 LINK: https://meteoraweb.com/intelligenza-artificiale/machine-learning-con-python-modelli-pronti-per-il-business-costi-controllati-e-deployment-reale

#computerVision #machineLearning #nLP #python #scikitLearn

Unlimited-OCR: One-shot Long-horizon OCR

https://github.com/baidu/Unlimited-OCR

#AI #OCR #ComputerVision

GitHub - baidu/Unlimited-OCR: Unlimited OCR Works: Welcome the Era of One-shot Long-horizon Parsing.

Unlimited OCR Works: Welcome the Era of One-shot Long-horizon Parsing. - baidu/Unlimited-OCR

GitHub