the « internet data lottery »: improvements in AI/ML coming from the path of largest web data availability rather than most scientifically sound direction modalities (text) win the lottery when they happen to be the widest available on the web even if sub-optimal for some tasks
MULTIMODAL AI
-
Magic3D: High-Resolution Text-to-3D Content Generation
By
–
Magic3D: High-Resolution Text-to-3D Content Creation Lin et al.: https://
arxiv.org/abs/2211.10440 #ArtificialIntelligence #DeepLearning #MachineLearning -
Generalist Vision-Language Models Emerging as Industry Standard
By
–
Authors propose a generalist model capable of handling major large-scale vision and vision-language tasks with competitive performance. These types of #AI models will become the industry norm over next 1 to 3 years.
-
Why CNNs Alone Cannot Achieve Level 5 Autonomous Vehicles
By
–
Plus there is a reason why AVs have not got to anywhere near Level 5 autonomy that Musk himself arrogantly predicted for 2020 – AI Vision (CNNs) alone is no where near enough. We need causal reasoning & learning on fly – CNNs alone won’t deliver truly advanced AVs @pierrepinna
-
LiDAR and Vision Fusion Enhances Autonomous Vehicle Safety
By
–
Ideally one should use both together which is why Tesla & Musk reportedly started to incorporate LiDAR to some extent. Vision (CNNs) alone will make mistakes that in real-world will be price costly. Combing both will reduce risk. https://
phase1vision.com/blog/how-lidar
-and-embedded-vision-increase-safety-in-autonomous-vehicles
… -
Runway AI Tools for Video and Image Editing
By
–
Runway – and Image editor (Free to try, $12/mo) @runwayml http://
runwayml.com editing, green screen, inpainting, and motion tracking. Change images with text descriptions, remove objects in videos, remove video backgrounds, expand images… and way more -
Cleanup Pictures: AI-Powered Image Object Removal Tool
By
–
Cleanup Pictures – Image Editor (Free +) by @cyrildiagne http://
cleanup.pictures Use to remove any unwanted text, logo, date stamp, or watermark, or part of an image. Super simple, fast tool. Free for 720p images -
Resemble AI enables custom voice model training
By
–
Resemble AI – Voice Cloning (free to try) @resembleai http://
resemble.ai Create a text-to-speech voice model using your *own voice* by training the model. This one is truly remarkable -
Mind Reading Technology, Brain-Computer Interfaces, and Latest Research Paper
By
–
Longer thread on "mind reading tech" and BCI and the latest paper to add to the mix. https://
mind-vis.github.io -
AI Computer Vision Detects Product Defects Perfectly
By
–
From Defect to Perfect! Impact of #AI #ComputerVision #ML #DeepLearning on Quality Great example @MarinerLLC @IntelIoT https://
insight.tech/industry/produ
ct-defect-detection-you-can-count-on-with-mariner-2?utm_source=twitter&utm_medium=organic&utm_campaign=2022-tdc-eaves
… #tech #DigitalTransformation @Shi4Tech #podcast #IntelPartner @insightdottech #ITInfluencer #Product @pierrepinna @SpirosMargaris