Joe Tubb’s Post

Architectural/AEC drawings hold immense value, but extracting insights has always been a challenge. We're sharing our thoughts on how AI-driven data extraction is revolutionizing the construction industry. Check out our latest Medium article on the potential of Drawing Retrieval Augmented Generation (DRAG) for AEC professionals! #AI #ConstructionTech #AEC AECforward.ai

A Complete Noobs Guide to Vector Search, Part 2 via #TowardsAI → https://bit.ly/3UsvPme

This is the fastest and funniest way to learn about generative AI. I Just came across this Generative IA DevApp platform called PartyRock where AI-generated apps can be built in a playground powered by Amazon Bedrock. https://lnkd.in/gfBnciEZ

Nice visualization to understand LLM https://bbycroft.net/llm #generativeai #llm

Multimodal models are here to stay. This video explains the underlying architectural approach of how this works. https://lnkd.in/gtSprxCJ

ARCHITEChTURES AI can import DWGs to model housing buildings for sunlight and daylight analysis, dynamic acoustics, and noise impact. By integrating applicable regulatory compliance, it enhances engineering designs, exports refined designs of floors, sections and facades (DWG), and provides documentation on costs, synopsis charts, or utilities estimation. (0:57 min video)

#polyglotnotebook #semantickernel #agenticworkflow AI Showoff: space-crawler using LLM agents Description: To monitor and bring down the hard disk space occupied by unnecessary files and folders. Design: Using agents having personas of File Analyzer, Directory Inspector and User Advisor

Transform your document handling processes with Veryfi Lens and experience unparalleled efficiency and accuracy. Ready to elevate your business? Learn more today! 🌐✨ #VeryfiLens #DataExtraction #OCR #Automation #BusinessEfficiency #AI #TechInnovation

EasyAnimate: A High-Performance Long Video Generation Method based on Transformer Architecture 🔔🔔🔔 EasyAnimate is an advanced video generation method leveraging transformer architecture, expanding the DiT framework for 2D images to 3D video by incorporating a motion module to capture temporal dynamics and ensure seamless transitions. This method adapts to various DiT baseline methods, generating videos with different styles, frame rates, and resolutions, and introduces slice VAE for long-duration video generation up to 144 frames. The holistic ecosystem for video production includes data pre-processing, VAE training, DiT model training, and end-to-end video inference. project: https://lnkd.in/dStn2_xP paper: https://lnkd.in/dnJmSzRG code: https://lnkd.in/dusm_KR7 #EasyAnimate #VideoGeneration #TransformerArchitecture #3DVideo #DiTFramework #MotionModule #TemporalDynamics #SliceVAE #VideoSynthesis #MachineLearning #AI #DeepLearning #ComputerVision #HighPerformanceAI #VideoProduction #DataProcessing #ModelTraining #VideoInference

Datature is excited to introduce our latest guide on YOLOv9 for advanced object detection tasks. This new article delves into the architecture of YOLOv9, its key differences, and how it compares against other predecessors. YOLOv9 introduces techniques such as Programmable Gradient Information (PGI) and Generalized Efficient Layer Aggregation Network (GELAN) to address information bottlenecks and enhance detection accuracy and efficiency in real-time object detection. Whether you're tackling complex medical images, enhancing security systems, or innovating in autonomous technology, this guide serves as a guide on what the hype is, and how you can fine-tune the model on your own dataset ✨ Read The Guide 👉 https://lnkd.in/gaEuwszd