Imagga

Many of you already know that at Imagga we are actively involved in making the online space a safer and more trustworthy place to be. Today, I'm excited to share a major milestone in this mission - our latest release of the Adult Content Detection model. This new version outperforms competitor solutions, offering up to 26% higher accuracy (benchmark dataset can be provided upon request)! This release marks the second iteration of our robust computer vision AI model, architected to detect explicit or suggestive nudity in online communities with machine- and/or user-generated content. Our model is available as part of our image recognition API stack and can be easily integrated as part of your solution to prevent such content from being uploaded and to safeguard the end-users. Our team is always available to assist you with the integration. Moreover, if you would like to test it, feel free send me a DM with your use-case and I will provide you with an exclusive access. #TrustAndSafety #NSFW #ContentModeration

Meet our new Adult Content Detection model. It ‘s designed for precise, real-time moderation of explicit videos and images. Keep your users safe and your platform compliant, and welcoming: https://lnkd.in/d5dXE-Un Our model outperforms leading competitors by up to 26% and delivers a substantial accuracy boost over our previous version (benchmark dataset available on request). This release marks the second iteration of our robust computer vision AI model, architected to detect explicit or suggestive nudity in online communities with both machine- and user-generated content. We’d love to hear from you – what types of content are hardest to manage, or where do existing detection models fall short? Let us know in the comments.

Happy to meet you 10:00 am tomorrow, on the symbolic date of Nevember 10, at Start Up, Bulgaria during my talk Turning Crisis into Opportunity: Scaling a Business Through Innovation. Will have the opportunity to share the story of Kelvin Health and how pioneering visual AI with Imagga long years ago enabled us to take an opportunity for global impact in healthcare at the very start of the COVID-19 pandemic. Co-founder Georgi Kostadinov, being one of the top AI architects worldwide, will be joining a panel on the recent dynamics of AI at noon the same day. Quite dynamic November for us, Start Up, Bulgaria being sandwiched right between MedTech World Malta and this year’s American Heart Association Scientific Sessions in Chicago, along with two EIT Health InnoStars Awards events at the very beginning and end of the month! #StartUpBulgaria #KelvinHealth #GlobalImpact

We're seeing more and more applications of image recognition in waste management. Read more in our latest blog post: https://lnkd.in/dC3Q5eNB #ai #imagerecognition #wastemanagement

AI adoption myths are everywhere. Here are a few common misunderstandings we’ve noticed from working with countless companies: - AI is plug-and-play - Little awareness of when a generic vs. custom models are needed - AI is cheap and easy to implement The reality? While generic models are budget-friendly and work very well for broad applications, they’re not a one-size-fits-all solution. Custom AI requires extensive data collection, specialized training, and significant costs At Imagga, we guide companies in identifying where AI fits best in their pipeline, providing clarity on what’s feasible now—and what still requires development.

Dating apps without content moderation are playing with fire. In the romance business, keeping users safe and engaged is everything. That’s why content moderation is a must-have. It can help: - Verify profile photos to keep profiles real - Block offensive content before it reaches users But moderation isn’t just about blocking - it’s about creating a space where people want to stay. With safety as the standard, users are more comfortable engaging, returning, and building connections. In a crowded market, authenticity and trust are your edge, and content moderation delivers both. Read more about the power of content moderation here: https://lnkd.in/dNZqK797 #ContentModeration #Dating

It's crucial for any ecosystem to bring bright people from the outside in order to flourish, exchange best practices, and build bridges that create business & collaboration opportunities. We are privileged that our city of Sofia just hosted such a high level event as #FutureForum by Webit.Foundation, Oct 23-24, passionately driven by Dr. Plamen Russev, Aniela Ruseff, and team. It was a pleasure to personally meet one of the Webit speakers, David Tan, ACEO of JTC Corporation, and his colleague Dany Yang in the office of Imagga and Kelvin Health. Pavel Andreev, Georgi Kostadinov, and I were amazed but the ambition and the efforts of Singapore towards sustainability and the opportunities this creates!

Yesterday, our 4th-year students in the Entrepreneurship course from the "Advertising" program at New Bulgarian University, which I have the pleasure of leading as a part-time university teacher, had the unique opportunity to learn from Georgi Kadrev, co-founder of Imagga and Kelvin Health. Georgi shared invaluable insights into the world of innovation and entrepreneurship, drawing on his extensive experience leading cutting-edge technology ventures. It was an inspiring session that left a lasting impression on our students. As a special highlight, we concluded the session with a demonstration of Kelvin Health’s groundbreaking thermal imaging technology—capturing a "thermal selfie" with the innovative tools developed by their team! Thank you, Georgi, for being part of this enriching experience and empowering the next generation of entrepreneurs! #Entrepreneurship #Innovation #TechInEducation #NBU #Advertising #KelvinHealth #Imagga

Synthetic data for content moderation: a breakthrough or a risk? 🔥 https://lnkd.in/dqe7iz-2 In the world of content moderation, where harmful content can spread like wildfire, we need AI models that are fast, accurate, and unbiased. But there's a problem: these models rely on massive, diverse datasets, which are nearly impossible to gather without running into privacy concerns or ethical issues. Real-world content is messy: - different cultures, contexts, and languages - privacy laws make data collection tricky - explicit or violent images are often off-limits One of the possible solutions is synthetic data. Generative AI can now create synthetic datasets to train these models without violating privacy or legal guidelines. But it’s not without its risks:  - bias if source data isn’t balanced - doesn’t fully capture real-world complexity In this blog post, Angel Spasov, Computer Vision Engineer & Researcher at Imagga, shares our team’s approach to using GenAI to augment and de-bias datasets: 🔥 https://lnkd.in/dqe7iz-2 #ContentModeration #AI #SyntheticData #GenAIData #Data #GenAI

AI-powered visual search is widely used in e-commerce improving customer experience and conversions. Here’s why it’s a game-changer for e-commerce: ✅ Improves product recommendations, giving customers exactly what they’re looking for ✅Offers personalized suggestions that align with customer preferences encourages more impulse buys ✅ Minimizes cart abandonment by suggesting relevant out-of-stock alternatives to keep shopping journeys uninterrupted Read more here: https://lnkd.in/eAZe_6JC Ready to take your e-commerce experience to the next level? Let’s chat! #AI #Ecommerce #VisualSearch #ConversionOptimization #Personalization #Imagga