- 15
- 3 701
Computer Vision Stream
เข้าร่วมเมื่อ 25 ก.ค. 2023
I help CV engineers
overcome deployment failures
by providing best practices/tools/ideas in MLOps/CVOps and Modern Software Engineering
overcome deployment failures
by providing best practices/tools/ideas in MLOps/CVOps and Modern Software Engineering
Video Summarization with VILA and LLama3.2 on Jetson
Learn how to rewrite Jetson services Video Summarization from GPT4 to llama3.2 hosted by OLLAMA.
we will find a classical video processing pipeline bug in the Nvidia sample explaining why VLM recognized blue school bust instead of yellow.
🔔 Don't forget to like, subscribe, and hit the notification bell to stay updated with our latest videos!
#Startups #Development #Technology #Innovation #Programming #C++ #Python
Code: github.com/Dorozhko-Anton/ComputerVisionStreams/tree/main/VLM_VideoSummarization
🔗 Connect with me:
Follow me on social media for more exciting updates and insights:
X: dorozhkoa
LinkedIn: www.linkedin.com/in/dorozhko-anton/
Book a consultation: dorozhko-anton.github.io
we will find a classical video processing pipeline bug in the Nvidia sample explaining why VLM recognized blue school bust instead of yellow.
🔔 Don't forget to like, subscribe, and hit the notification bell to stay updated with our latest videos!
#Startups #Development #Technology #Innovation #Programming #C++ #Python
Code: github.com/Dorozhko-Anton/ComputerVisionStreams/tree/main/VLM_VideoSummarization
🔗 Connect with me:
Follow me on social media for more exciting updates and insights:
X: dorozhkoa
LinkedIn: www.linkedin.com/in/dorozhko-anton/
Book a consultation: dorozhko-anton.github.io
มุมมอง: 103
วีดีโอ
Object detection on Jetson without training data (zero-shot detection on Jetson)
มุมมอง 80วันที่ผ่านมา
Zero-shot detection with NanoOWL on Jetson devices. In this video, we cover a video processing pipeline to deploy open-vocabulary object detection that can handle even objects for which you don't have training data. 🔔 Don't forget to like, subscribe, and hit the notification bell to stay updated with our latest videos! #Startups #Development #Technology #Innovation #Programming #C #Python Code:...
You Won't Believe How Easy RTSP Streams Are with NVIDIA NVstreamer!
มุมมอง 47วันที่ผ่านมา
Generate RTSP streams from video files to test your video processing pipelines or setup a live demo for your clients. 🔔 Don't forget to like, subscribe, and hit the notification bell to stay updated with our latest videos! #Startups #Development #Technology #Innovation #Programming #C #Python Code: github.com/Dorozhko-Anton/ComputerVisionStreams/tree/main/CreateRTSPStream 🔗 Connect with me: Fol...
Avoid DEPENDENCY HELL and save time with Semantic Versioning 2.0.0
มุมมอง 314 หลายเดือนก่อน
In this video, we explore Semantic Versioning (SemVer)-a systematic approach to versioning software that helps developers manage dependencies, ensure compatibility, and communicate changes effectively. We’ll break down the MAJOR.MINOR.PATCH format, explain how it signals the impact of updates, and demonstrate how SemVer simplifies release planning and automation in software development. Whether...
DeepStream MultiObject Tracking Evaluation
มุมมอง 1734 หลายเดือนก่อน
In this video, we demonstrate how to extract tracking metadata in KITTI format, convert it to MOT format, and apply tracking evaluation to compute key multi-object tracking (MOT) metrics. Starting with a DeepStream pipeline that processes a video file, runs object detection, and utilizes an NVIDIA tracker, we walk through the process of adding a probe to dump tracking information. We guide you ...
Enhancing Performance: Measuring DeepStream Latency
มุมมอง 965 หลายเดือนก่อน
In this video, we explore how to measure elements latency within a DeepStream pipeline. You'll learn the importance of latency monitoring in real-time AI applications and how to utilize Gstreamer tracing capabilities to accurately track and optimize your pipeline's performance. Whether you're working on object detection, video analytics, or any other AI-powered solution, understanding latency i...
Nvidia's DeepStream 7.0 Service Maker : Simplifying computer vision app development.
มุมมอง 3585 หลายเดือนก่อน
Are you a DeepStream developer looking to leverage GStreamer but find the learning curve too steep? We've got fantastic news for you! NVIDIA is excited to introduce a groundbreaking feature in the DeepStream technology suite: DeepStream Service Maker. This innovative addition is designed to simplify the development process by abstracting the complexities of GStreamer, allowing you to build C ob...
Best 35$ investment for every software engineer!
มุมมอง 208ปีที่แล้ว
Modern Software Engineering Book: www.amazon.com/Modern-Software-Engineering-Discipline-Development/dp/0137314914 Also highly recommend David Farley's YT channel: Continuous Delivery @ContinuousDelivery 🔗 Connect with me: Follow me on social media for more exciting updates and insights: X: dorozhkoa LinkedIn: www.linkedin.com/in/dorozhko-anton/
(Part 4/5) 50 Computer Vision Applications for Smart City
มุมมอง 51ปีที่แล้ว
🔍 What's Inside: In the first part of this series, we unveil the next 10 out of 50 remarkable computer vision applications shaping the cities of tomorrow. Discover how AI-powered cameras, sensors, and data analysis drive innovation across various urban life aspects. From real-time traffic optimization to intelligent waste management, each application brings us one step closer to the cities of o...
50 Comupter Vision Applications for Smart City (Part 3)
มุมมอง 36ปีที่แล้ว
🔍 What's Inside: In the first part of this series, we unveil the next 13 out of 50 remarkable computer vision applications shaping the cities of tomorrow. Discover how AI-powered cameras, sensors and data analysis drive innovation across various urban life aspects. From real-time traffic optimization to intelligent waste management, each application brings us one step closer to the cities of ou...
50 Computer Vision Applications for Smart City (part 2)
มุมมอง 69ปีที่แล้ว
🔍 What's Inside: In the first part of this series, we unveil the next 12 out of 50 remarkable computer vision applications shaping the cities of tomorrow. Discover how AI-powered cameras, sensors and data analysis drive innovation across various urban life aspects. From real-time traffic optimization to intelligent waste management, each application brings us one step closer to the cities of ou...
50 Computer Vision Applications for Smart City (Part 1)
มุมมอง 157ปีที่แล้ว
🔍 What's Inside: In the first part of this series, we unveil the first 5 out of 50 remarkable computer vision applications that are shaping the cities of tomorrow. Discover how AI-powered cameras, sensors, and data analysis are driving innovation across various aspects of urban life. From real-time traffic optimization to intelligent waste management, each application brings us one step closer ...
Consume any video source in OpenCV with GStreamer
มุมมอง 1.5Kปีที่แล้ว
Learn how to retrive and preprocess images from any cameras without loosing a comfort of staying in the Python dev environment of OpenCV by leveraging the power of GStreamer pipelines. Code: github.com/Dorozhko-Anton/ComputerVisionStreams/tree/main/Opencv+Gstreamer Gst source plugins: github.com/AravisProject/aravis github.com/basler/gst-plugin-pylon
Level Up Your Computer Vision Skills with GStreamer: An Introductory Tutorial
มุมมอง 572ปีที่แล้ว
Learn the basic abstraction of GStreamer library - multimedia framework that will help you bring you Computer Vision solutions to production by simplifying retrival of video and images from different sources (including RTSP streams and specialized industrial cameras) Code: github.com/Dorozhko-Anton/ComputerVisionStreams/tree/main/Introduction_Gstreamer
Master the Art of Testing Computer Vision with Real-World Streams
มุมมอง 194ปีที่แล้ว
How to use CCTV cameras that are accessible online to benchmark your #deepstream #computervision pipelines? Code: github.com/Dorozhko-Anton/ComputerVisionStreams/tree/main/Deepstream_vs_TH-cam
Hi Anton, the audio seems to be distorted, this happens when you speedup the video during editing. Instead you can use ffmpeg on the final rendered video to avoid such issues. Here is the command that worked for me. This speedups by 1.3x ffmpeg -i input.mp4 -vf "setpts=(PTS-STARTPTS)/1.3" -crf 18 -af atempo=1.3 speedup.mp4
Audio is not clear - at least to me
Thanks for the comment, will improve in the next video.
dont stop, this is good stuff
Thank you so much for the encouragement, Amir! Will definitely share more interesting videos about computer vision.
In my OpenCV built using python Gstreamer is "NO" and I do have Gstreamer installed. How I can gstreamer building in opencv??
You can try to rebuild opencv from source and activate gstreamer support with cmake flags
@@ComputerVisionStream Did this but still it says no to gstreamer
Great video! Keep going 👍
Thank you! 👍
Amazing!
Thanks, glad that you liked it!
Keep doing what you are doing bro!
Thanks, will do!
Very interesting ! Thank you
Glad you liked it!
Nice insight into CV in practice 👍
Thanks! There will be more soon.😁