๐ What if AI could detect and count boats directly from drone footage? ๐
Experimenting with @roboflow + SAM 3 to turn aerial imagery into actionable insights through Computer Vision! ๐ค
๐ฅ Video credit: Instagram my_shutter_life
#Roboflow#AI#ComputerVision#YOLO
Exactly 5 years ago today, Windies Cricket shared my video and made this Maroon fanโs dream moment unforgettable.๐ฅนโค๏ธ๐ Once a Maroon, always a Maroon.
๐ From Hugging Face to NVIDIA Jetson in One Step!
Embedl opens its platform to the public with 1,000+ pre-optimized AI models, one-command deployment, and browser-based testing on real Jetson devices. ๐ฅ
#Embedl#EdgeAI#NVIDIAJetson#ComputerVision#AI#Robotics
๐๏ธ ๐๐ป๐๐ฒ๐ฟ๐ป๐ฎ๐๐ถ๐ผ๐ป๐ฎ๐น ๐๐ฟ๐ฒ๐ฎ๐๐ ๐๐ฎ๐ป๐ฐ๐ฒ๐ฟ ๐๐๐ฎ๐ฟ๐ฒ๐ป๐ฒ๐๐ ๐๐ฎ๐ | ๐๐ ๐ฝ๐น๐ผ๐ฟ๐ถ๐ป๐ด ๐ ๐ฒ๐ฑ๐ถ๐ฐ๐ฎ๐น ๐๐ ๐
๐ฌ ๐๐ ๐ฝ๐น๐ผ๐ฟ๐ถ๐ป๐ด ๐๐ต๐ฒ ๐๐ฟ๐ฒ๐ฎ๐๐ ๐จ๐น๐๐ฟ๐ฎ๐๐ผ๐๐ป๐ฑ ๐๐บ๐ฎ๐ด๐ฒ๐ ๐ฑ๐ฎ๐๐ฎ๐๐ฒ๐ ๐ผ๐ป @ultralytics ๐ฃ๐น๐ฎ๐๐ณ๐ผ๐ฟ๐บ.
๐ Dataset highlights:
โข 778 ultrasound images
โข 3 classification classes: Benign, Malignant, and Normal
โข 622 training images, 156 validation images
โข 47.9 MB dataset size
๐ง These ultrasound images can help researchers explore how AI models learn patterns in breast tissue and investigate potential applications in breast-lesion classification.
๐ก Medical AI has exciting potential, but clinical validation and expert interpretation remain essential.
๐ ๐๐ฒ๐ $๐ฏ๐ฌ ๐ถ๐ป ๐จ๐น๐๐ฟ๐ฎ๐น๐๐๐ถ๐ฐ๐ ๐ฃ๐น๐ฎ๐๐ณ๐ผ๐ฟ๐บ ๐ฐ๐ฟ๐ฒ๐ฑ๐ถ๐๐ ๐๐๐ถ๐ป๐ด ๐บ๐ ๐ฐ๐ผ๐ฑ๐ฒ: ๐๐ข๐๐-๐จ๐๐ง ๐
๐ https://t.co/w80MX4fuKm
๐Dataset --> https://t.co/vOOGuAmwmd
๐ฌ How do you think AI can contribute to the future of breast cancer screening?
#BreastCancerAwareness #MedicalAI #ComputerVision #Ultralytics #HealthcareAI
๐ What if AI could see an entire city through a droneโs eyes? ๐๐ค
โ๏ธ ๐ฆ๐๐ป๐ฑ๐ฎ๐ ๐๐ ๐๐ ๐ฝ๐ฒ๐ฟ๐ถ๐บ๐ฒ๐ป๐ | ๐ฉ๐ถ๐๐๐ฟ๐ผ๐ป๐ฒ ๐๐ฎ๐๐ฎ๐๐ฒ๐
Testing object detection on real-world aerial footage, identifying cars ๐ and pedestrians ๐ถ using Computer Vision.
Small experiments. Real-world AI. Endless possibilities. ๐
๐ฌ What would YOU build with drone-based AI? Drop your ideas below! ๐
๐ฅ ๐๐ผ๐น๐น๐ผ๐ ๐บ๐ฒ ๐ณ๐ผ๐ฟ ๐บ๐ผ๐ฟ๐ฒ ๐๐ผ๐บ๐ฝ๐๐๐ฒ๐ฟ ๐ฉ๐ถ๐๐ถ๐ผ๐ป ๐ฝ๐ฟ๐ผ๐ท๐ฒ๐ฐ๐๐ ๐ฎ๐ป๐ฑ ๐ต๐ฎ๐ป๐ฑ๐-๐ผ๐ป ๐๐ ๐ถ๐ป๐ป๐ผ๐๐ฎ๐๐ถ๐ผ๐ป๐.
#ComputerVision #VisDrone #ArtificialIntelligence #DeepLearning
๐คฏ What if AI could count every bottle cap automatically on a production line? ๐ญ๐
Using @ultralytics YOLO + Computer Vision, we can detect and count bottle caps to support smarter manufacturing and quality control ๐
๐ก Real-world applications:
๐ฆ Automated counting
โ๏ธ Production monitoring
๐ฏ Quality inspection
๐ค Industrial automation
๐ Get $30 in Ultralytics Platform credits with my code: JOEL-ULT ๐
๐ https://t.co/w80MX4fuKm
๐ฌ Where would you use AI-powered counting?
Follow me for more YOLO & Computer Vision projects ๐ฅ
#Ultralytics #YOLO #ComputerVision #AI #ObjectDetection #IndustrialAutomation #Manufacturing
๐ From @LightOnIO OCR-2-1B to LightOnOCR-3!
A few months ago, I tested LightOnOCR-2-1B. Now, I'm exploring LightOnOCR-3๐ค๐
โก Faster transcription
๐ Document element detection
๐ผ๏ธ Image descriptions
๐ Chart data extraction
OCR is evolving beyond text extraction toward intelligent Document AI.
๐ฅ Demo: https://t.co/wXe9z6Hna9
๐ LinkedIn post: https://t.co/GB9zzCVvy9
๐ Previous experiment: https://t.co/Fn6kCx2yaq
๐ฌ Have you tried VLM-powered OCR yet?
Follow for more Computer Vision & AI experiments. Open to collaborations and meetups ๐ค
#LightOnOCR #DocumentAI #OCR #VLM #ComputerVision #AI
Select images โ run plugin โ bounding boxes. โก๏ธ
LightlyTrain brings model inference into your data workflow - no labeling scripts, no handoffs.
#AutoLabeling#ObjectDetection
๐คฏ What if AI could count every item on a conveyor belt automatically?
๐ญ YOLO-powered Computer Vision enables real-time object detection and counting for smarter manufacturing! ๐ฆโก
๐ Get $30 in @ultralytics credits with code JOEL-ULT ๐ https://t.co/w80MX4fuKm
Follow for more AI & Computer Vision๐
#YOLO #Ultralytics #ComputerVision #AI #Automation
๐ Oil spill segmentation is more than drawing masks โ itโs turning pixels into actionable insights.
๐ค YOLO-powered AI can help map affected areas, support monitoring, and guide environmental response. ๐ฐ๏ธ๐
๐ Get $30 Ultralytics credit with code JOEL-ULT ๐ https://t.co/w80MX4fuKm
#YOLO #ComputerVision #AI #ImageSegmentation #ClimateTech
๐ค Hugging Face is where AI meets open source!
๐ค Models | ๐ Datasets | โก Spaces | ๐ Community
From LLMs and Computer Vision to VLMs and AI agents, explore, build, share and innovate together ๐
๐ค Open to AI collaborations.
#HuggingFace#AI#OpenSource#MachineLearning #ComputerVision
๐ Embedl optimizes AI for edge devices with quantization, pruning, NAS & knowledge distillation. โก
๐น Embedl Hub ๐น Model Optimization SDK ๐น Ready-to-run AI models ๐น CPU, GPU, FPGA & ASIC support
๐ค Open to Edge AI collaborations
#Embedl #EdgeAI#AI #MLOps #ComputerVision #CVOps
๐ @TheRealDagsHub is more than Git for ML.
It brings code + datasets + models + experiments + collaboration into one workflow. ๐ค
๐ฆ Data & dataset versioning
๐ Git + DVC
๐งช Experiment tracking
๐ค Model versioning
๐ฅ Team collaboration
๐ Reproducible ML workflows
๐ Research โ Production
The idea is simple:
Data โ Version โ Train โ Track โ Evaluate โ Collaborate โ Deploy
Perfect for ML, Computer Vision, MLOps & open-source projects.
๐ค Open to collaboration.
๐ Follow me for more AI, ML & MLOps content.
#DagsHub #MLOps #AI #MachineLearning #ComputerVision #OpenSource #DVC
๐ OpenMMLab is a powerful open-source ecosystem for Computer Vision โ from detection & segmentation to pose estimation, 3D vision, action recognition, OCR & deployment. ๐ค๐๏ธ
๐ค Open to AI & Computer Vision collaborations!
๐ Follow me for more AI, Computer Vision & CVOps content.
#OpenMMLab #ComputerVision #AI #DeepLearning #OpenSource #MedicalImaging #CVOps