Atomix’s Post

first time using yolov9 I want to compare it with yolov8 on a complicated project can't wait to see results #objectdetection #computervision #innovation

🚀 We've just launched Scandit Express 3.0 🚀 This upgrade of Scandit Express is now more advanced, more compatible and more customizable. 👉 Scan and count multiple barcodes simultaneously with the Inventory Count feature, which offers you AR feedback to get the job done 👉 More keyboard customization options such as languages and sizes 👉 Compatibility with apps that don’t include a text field 👉 Automatically parse barcode data and populate CSV files And all of this is accessible in just a few minutes, thanks to Scandit Express' no-code integration. Excited about the possibilities? Get all the details about Scandit Express 3.0 here 👉 https://lnkd.in/ddhM5rCi #BarcodeScanning #SmartDataCapture #AugmentedReality #Scanning

💡 #SingularXYZ Weekly Tech Chat | How to Use Receiver IMU as a #Compass for Stakeout? Have you ever encountered difficulties while staking out due to a malfunctioning controller or phone compass? In this case, it is difficult to find the desired point. However, the latest SingularPad software supports stakeout using the receiver's IMU as a compass, ensuring accurate navigation even if the PDA compass fails. Discover more in the #blog 👉 https://lnkd.in/gsQQk_v9 #GPS #GNSS #geospatial #geodetic #surveyor #surveyinglife #surveyorlife #surveying #techblog #Galileo #rtk #stakeout #IMU

LLMs are literally the most unreliable technology of all time (followed by **ing bluetooth) After an absurd amount of trial and error at Builder.io, we've internally created a set of rules for making LLMs considerably more reliable Here's a brain dump of our secrets: https://lnkd.in/g_HShsMK

The post below aligns to my statement from last week: LLM-based AI architectures are complicated, and the location where an AI-output is optimal is narrow (i.e. could be just 'at the end of a process'). There's also the fact that everything that is AI is LLM - meaning all the OTHER areas of Machine Learning are either ignored, neglected or not mentioned (as they don't fit in the LLM-AI narrative).

The improvements in accuracy are very impressive. Even more impressive are the latency values. I can't wait to see the real-time performance of this model. For those who want to explore it in detail, here is the GitHub link: https://lnkd.in/gyUyXNUa

This video explores Wireless InSite’s basic antenna functionality. Users will be introduced to several built-in antenna types, how to import their own user-defined antenna patterns, and Wireless InSite’s built-in functionality for defining and visualizing antenna patterns within the software. #wireless #propagation #antenna

Yes, we are really talking to you! 😎 At Amperecloud we often develop new products together with partners who are already using our software. Well, from now own everyone can submit ideas for features that we are not having (yet). 💙 Our new product ideas board is publicly available and over 20 ideas have already been submitted. So what are you waiting for? (link in the comments)

Giving users a choice between faster processing speeds and lowered costs is one of the new features found in our testnet. Read about all our latest features here 👇 https://lnkd.in/dQqUvsrr 

The READ button gives the incendo user the ability to READ a pre-programmed NFC tag and enable the system to "clone" it. It is a very useful way to quickly setup the system to program a certain type of model or batch based DPPs from the system. The READ function reads the NFC tags and then shows the result what Product ID it found on the chip.