Advex specializes in leveraging generative AI and synthetic data to enhance computer vision systems by reducing data collection times and improving model accuracy. The company self-describes its services as crucial for automating data generation.

Advex

Advex AI  | 1,536 followers on LinkedIn. Solve Data. Solve AI. | Advex AI helps the world&#39;s largest manufacturers reduce costs and improve efficiency by providing highly reliable computer vision models in hours. 

Our models are self-healing and robust to environmental changes, so they stay reliable over time without manual re-training. To do this, we leverage proprietary technology that finds gaps in your datasets and fixes them with synthetic data.

Hiring at Advex AI | ex-Waymo, ex-Twitter

Catherine Li

Founding Engineer at AdvexAi; ex-Machine Learning Engineer at Scale.AI

Benjamin Pikus

Building and Learning | 8VC Fellow | KP Fellow | BVP Fellow | WashU CS

David Anyaeche

Co-founder / CTO Advex AI | Solving OOD Generalization

Qasim Wani

Advex AI

Thanks for the informative post Gajen Kandiah !

We're thrilled to announce that our CEO Pedro Pachuca will be presenting on the TechCrunch Disrupt main stage today as part of the 2024 Startup Battlefield Top 20! 🔊 🚀 Immense gratitude to our incredible team and investors who have helped make this milestone possible. #TechCrunch #StartupBattlefield #AI #Innovation "

One week until #StartupBattlefield 🚀 🚀 🚀 Visit Booth H9 at TechCrunch Disrupt, Oct 28-30 in San Francisco and come see our breakthrough demo: an AI vision system supercharged by synthetic data that spots critical manufacturing defects with unprecedented accuracy. #TechCrunchDisrupt2024 #StartupBattlefield #AIInnovation #ComputerVision Nicholas Blake Benjamin Pikus Qasim Wani Catherine Li Pedro Pachuca Brandon Trabucco Pear VC Construct Capital Emerson Collective John Reuter

Kicking off our series on interesting ideas from #ECCV2024 ! Over the next few posts, we'll highlight cutting-edge research happening in computer vision today. Let's dive into our first topic! David Bau presented an interesting approach to model unlearning for diffusion models ( https://lnkd.in/gc7MZV6q ). Here's the breakdown: 🔑 Key Innovation: Performing model editing on large diffusion models without having to retrain the entire model. 👩💻How it works: 1. Reduces the probability of generating an image according to its likelihood given only a single textual description of the concept to be removed. 2. Utilizes Tweedie’s formula, a concept first discovered in 1956 by Maurice Robbins and later popularized by the renowned statistician Bradley Efron , offers a simple Bayes approach for correcting selection bias. 3. Utilizing this trick, the diffusion model is able to move its data distribution to minimize the generation probability of the unwanted concept without having to retrain the entire model on filtered data. 🚀 Limitations: Erasing large concepts such as entire object classes doesn’t work well due to how entangled the features are across the weights of the denoising model. 📸 Why this matters for effective synthetic data generation: Unlearning for diffusion models can help focus synthetic image generation towards concepts you want and away from concepts you don’t want, enabling you to generate higher quality synthetic images without re-training the model on a large amount of data. #ECCV2024 #SolveDataSolveAI #AdvexAI

VISION 2024: Wrapping Up and Spotlighting Our Neighbors 🤝🔍 What an incredible week at VISION 2024 in Stuttgart! The excitement around AI in machine vision and our synthetic data platform was palpable. We're inspired by the conversations on edge computing, advanced sensors, and cross-industry collaboration. Let's shine a spotlight on some of our fantastic neighbors: Baumer Group GmbH (Booth 8D40) ☕🍫 Great coffee, chocolate, and tech! Baumer impressed with their camera portfolio for non-visible wavelength applications and GigE Vision 3.0 with up to 100 GigE over RDMA. New Imaging Technologies (NIT) 👁️ NIT pushed boundaries with their cutting-edge SWIR cameras, expanding machine perception beyond visible light. Nortech Systems, Inc. 🔌 Nortech's "Intercon 1" machine vision products showcased the intricate ecosystem supporting machine vision advancements. IMAGO Technologies GmbH 🖥️🤖 IMAGO's intelligent, programmable cameras and AI-ready solutions aligned perfectly with the industry's future direction. We're grateful to be part of such an innovative community. Missed us at VISION 2024? Reach out for a demo of our synthetic data platform! Here's to pushing the boundaries of machine vision together. Until next time, Stuttgart! #VISION2024 #MachineVision #SyntheticData #AIInnovation #AdvexAI

Want to work on some of the hardest problems in AI today? Come join us as we #SolveData to #SolveAI .

👁️ Exploring the future of machine vision at VISION 2024 in Stuttgart! 🤖 AI is the cornerstone of next-gen machine vision, and our synthetic data is fueling its rapid evolution. But breakthrough AI requires a holistic approach. We're here to explore cutting-edge sensors, lighting, and on-the-edge computing solutions that complement our synthetic data platform. 🫂 Join the AI revolution in Hall 8 (Alfred-Kärcher), Stand 8C38. Let's push the boundaries of what machines can see and understand. 🐖 Fun Stuttgart fact: Stuttgart is home to the world's largest pig museum. Wonder if they're using vision to determine which pig is the most photogenic? #VISION2024 #SolveDataSolveAI #MachineVision #Pigs

#ECCV Day 2 ✅ The people have spoken. The main challenge to achieving production-grade AI model is collecting data. Come check us out at booth #40 to see how Advex solves the data problem for the most important industry in the world! #SolveDataSolveAI

ADVE

AI:ADVEX

Advex