Roboflow

Incredibly excited and humbled to share I’ve joined Roboflow as head of sales to lead our next phase of revenue and GTM scaling. Excited because computer vision is a platform-level shift akin to mobile and the web itself. Roboflow builds the ecosystem to realize the potential of visual AI/ML: open source tools, community, and a hosted platform to create + deploy computer vision AI workflows. Humbled by the quality of people I spent the week with in NYC for my first week on board. Everyone in the company is laser focused on making customers/users successful. We work quickly, with pride, and autonomy- completely off the charts group of people. We’ve got the ingredients to build something very special and long lasting. Let's get after it Joseph Nelson, Brad Dwyer, Trevor Lynn, Kate Wagner, Jacob Witt and the whole Roboflow crew!

Three new tutorial notebooks are out! 🥵 🥵 🥵 (great models under permissive licenses) Our collection of notebooks is growing. In just the last few weeks, we've added three tutorials covering two models - Florence-2 and RT-DETR. - Florence-2 - lightweight vision-language model open-sourced by Microsoft under the MIT license. The model demonstrates strong zero-shot and fine-tuning capabilities across tasks such as captioning, object detection, grounding, and segmentation. Despite its small size, it achieves results on par with models many times larger, like Kosmos-2. - RT-DETR - short for "Real-Time DEtection TRansformer", is a computer vision model developed by Peking University and Baidu, Inc. In their paper, "DETRs Beat YOLOs on Real-time Object Detection" the authors claim that RT-DETR can outperform YOLO models in object detection, both in terms of speed and accuracy. The model has been released under the Apache 2.0 license, making it a great option, especially for enterprise projects. I've written two blog posts about Florence-2 - one discussing the model's capabilities across various computer vision tasks and the other on fine-tuning Florence-2 on custom object detection. I've also published a YouTube tutorial covering this model. The RT-DETR fine-tuning blog post will be out today. Links to all these resources are in the description below. 👇🏻 ⮑ 🔗 notebooks repository: https://lnkd.in/deQexCeS #tutorial #computervision #objectdetection #multimodal #artificialintelligence

🔴 live: run different vision tasks with Florence-2 We will discuss object detection, different levels of image captioning, phrase grounding, open vocabulary detection, region proposal, OCR, and more. I'm going live now! ⮑ 🔗 YouTube stream: https://lnkd.in/dT33Zz_M #computervision #objectdetection #transformers #multimodalai hashtag #opensource

📜 50% of Supervision features on one Cheatsheet 🌳 https://lnkd.in/dkJmEQmV Supervision is growing, and recalling all the method definitions is getting harder. Even for someone working at Roboflow. So I compiled this cheatsheet, featuring the most commonly used features! let's see if it helps you as much as it helps me 😉 Thanks to Bhavay Malhotra for helping me out!

How do you connect this old typewriter to a computer using just a cheap camera? 🤔 One of the first things I'm tackling at Roboflow is a "Visionary Project," where I'm diving into the tools and models our customers use to get their computer vision projects live. It's like diving headfirst into our own product, right? After my first week at our SF office, I came back with a typewriter! This relic from the 40s still types like a champ. Now I'm on a mission: figuring out the quickest way to hook it up to a computer so I can type up some cool stuff with this beauty. I've got a few ideas, but I'm curious: how would you go about solving this problem with #computervision?

Fine-tune Florence-2 on custom detection dataset - YouTube tutorial 🔥 🔥 🔥 Finally, after almost a week of work, I released my tutorial about fine-tuning Florence-2, a computer vision model released by Microsoft, on a custom object detection dataset. 😩 I really hope you'll like it. I'm so tired, haha Topics covered: - Running the pre-trained model with various vision tasks - Configuring LoRA for efficient fine-tuning - Deep dive into the Florence-2 compatible dataset format - Training and benchmarking the fine-tuned model - Comparative analysis: Florence-2 vs. YOLOv8 for fine-tuning - Florence-2 vs. top vision models ⮑ 🔗 YouTube video: https://lnkd.in/dGxHfAUf Links to the paper, code, and blog post are in the comments below. 👇🏻 #computervision #objectdetection #transformers #multimodalai #opensource