Skit.ai’s Post

[ SDXL - Cyborg and Girl ] While creating a futuristic background, I tried using SDXL. Rather than contrasting machine and human, I wanted to depict characters that didn't get along with each other. The robot is WHITE, and the girl is ANN BLACK. 🙄 It was created based on the Prompt, without any particular ControlNet. There were many attempts because the formation didn't go as well as I thought. #stablediffusion #aiart #sdxl #white #black #soylab #stablediffusionkorea

𝐒𝐰𝐚𝐩 𝐚𝐧𝐲 𝐟𝐚𝐜𝐞 𝐰𝐢𝐭𝐡 𝐞𝐚𝐬𝐞. using: 1️⃣ 𝘙𝘰𝘰𝘱 2️⃣ 𝘚𝘵𝘢𝘣𝘭𝘦 𝘋𝘪𝘧𝘧𝘶𝘴𝘪𝘰𝘯 3️⃣ 𝘈𝘶𝘵𝘰𝘮𝘢𝘵𝘪𝘤 1111 4️⃣ 𝘖𝘯𝘦 𝘪𝘮𝘢𝘨𝘦 𝘰𝘧 𝘧𝘢𝘤𝘦 --- These tools are actually becoming exceptionally good at face-swapping with just one image, using Roop and stable diffusion --- 👋 Hi it's sam , If you liked this, connect with me on LinkedIn for daily insights on #machinelearning #genai #roop #stablediffusion #aiimages

Excited to share my recent computer vision project 'Salamatk': a Driver Behavior Monitoring System built with YOLOv8! Salamatk focuses on real-time detection of potentially dangerous behaviors while driving, such as: 📱 Using a phone 🍔 Eating or drinking 😴 Yawning (indicating fatigue) 🚬 Smoking The system not only identifies these behaviors with high accuracy but also takes immediate action for certain hazardous activities. For instances of phone use or signs of driver fatigue like yawning, it triggers an alarm sound to alert the driver, fostering a safer driving environment. Developing "Salamatk" involved three main steps: 1- Collecting and labeling data for accuracy. 2- Adjusting YOLOv8 to improve speed and detection. 3- Implementing a system for instant behavior analysis and safety intervention. Special thanks to my teammates Khaled Al-shehri, Bader Alomari and Mutaz Altalhi for their hard work and dedication. For a closer look at our project, including code and documentation, check out our GitHub repository: https://lnkd.in/dbnN4iQd I'd love to hear your thoughts, feedback, or any insights on similar technologies. #ComputerVision #YOLOv8 #ArtificialIntelligence #AI #RoadSafety

🚨 ThirdEye Unveils DeepfakeAI for Defense Use! 🚨 🔍 Worried about deepfakes wreaking havoc online? ThirdEye's got your back! Our new DeepfakeAI software is here to save the day, helping you verify your identity and remove harmful deepfakes effortlessly. 💰 With deepfakes causing over $5B+ in damages, it's time to take action! Don't fall victim to misinformation - arm yourself with ThirdEye's cutting-edge technology. Plus, our tool can even detect AI alterations in text, audio, and video files! Contact [email protected] or visit https://thirdeyegen.com to learn more! #DeepfakeAI #ArtificialIntelligence #MixedReality #ThirdEyeGen #smartglasses #TechNews

Here is a TED-Talk about how AI will move from the digital world to the real world. Kicker is that MANUS™ is right between both of these worlds. Serving as a data gathering solution to teach AI how to perform tasks and engage with the real world. At 10:20 you see our previous generation Prime X gloves.

You might have seen those wobbly, distorted letters and numbers you have to type in to prove you're not a robot when signing up for things online. They have a cool name: CAPTCHA. It means "Completely Automated Public Turing test to tell Computers and Humans Apart."Scientists came up with it to tell real people from pesky bots. #rfsfacts #digitalpuzzles #websecurity #humanvsbot #CAPTCHAfun #onlineprivacy #digitalknowledge #internetsecurity #funfact #learnsomethingnew #webtrivia #factoftheday #onlineprotection #knowledgebite

The rise of vaping among students is a growing concern. Claro Enterprise Solutions has a solution w/ AI Video Analytics's Vaping Detection. Protect your students wellbeing with our AI-powered technology. #SchoolSafety #VapingDetection #StudentSafety

Quick rundown of testing ChatGPT-4o with images of guns, gun-shaped objects, and toy guns. Based on these tests, I think OpenAI the most advanced weapon detection technology ever created and this is just the start!

#Adversarial attacks on #LLMs: Human #RedTeaming #TrustEverybodyButCutTheCards https://lnkd.in/ds3P7ECu introduce #Bot-Adversarial Dialogue, a human-and-model-in-the-loop framework for enhancing conversational #AIsafety. #Crowdworkers converse with #chatbots to elicit unsafe/offensive responses, categorized by severity. A verification task identifies #offensivelanguage types, involving humans in both collecting and #labeling adversarial examples for #safety and offensiveness type. #TrustEverybodyButCutTheCards

Why Can’t Robots Click the “I’m Not a Robot” Box? 🤖🚫 Ever wondered why bots can’t just click that “I’m not a robot” box? Here’s the scoop: 🖱️ How You Click Matters: It’s not just about clicking the box. Bots are too fast and efficient, while humans are more random and slower with their movements. 🧩 Visual Puzzles: If you click too quickly, you get visual puzzles. Bots struggle with these because they can’t easily pick out objects in cluttered images. 🔍 Invisible Methods: Some websites use invisible CAPTCHA methods to check your browsing history, mouse movements, and behavior to determine if you’re human. Next time you click that box, remember it’s all about proving your humanity through your unique, human-like inefficiency! #TechTidbits #OnlineSecurity #reCAPTCHA