jfhb-04

Joined 3 years ago

Karma
24
Submissions
16
Member since
Oct 12
Building futvis - a platform for every soccer player to generate video soccer analytics out of their match videos

X: @joseherazo04

View on Hacker News

Recent submissions

Comment · 3 years ago
Didn't think about this! Thanks for this idea! It sounds like it could be extremely useful for referees.
Comment · 3 years ago
I'm focusing on amateur leagues and amateur teams.

In fact, I'm an amateur player and I'm in a team that plays every single week. We collect the videos on our own and process them to have our analysis and make decisions. It's really low cost since we use our smartphones and don't need the analysis in real-time.

By now we're building a platform where you upload your videos and access your analysis, in contrast to others, we are not focusing on video recording.

Comment · 3 years ago
I understand. Since we're still in the process of building the platform, this is the only way to generate insights for you currently. However, in a few weeks, the platform will be completed, allowing you to use it on your own while maintaining ownership of your videos. I'll send you an email when it's done!
Comment · 3 years ago
By now, we're only capable of processing videos with consistent camera angle throughout the video, I believe trace cameras change based on players' positions, right?

I'm interested in adjusting our systems and make it work with trace cameras.

If you want send me a message and we can work to generate some data from your video. It'll help me a lot to improve Futvis.

[email protected]

Comment · 3 years ago
I'm so sorry for this, it was my bad. On the website, there is a link to the demo people can try but it is not exactly the link I posted.

This is the link: https://futvis.streamlit.app/. Sorry for not having it accessible to anyone. Unfortunately, it is a button inside the website and I misunderstood the rules.

Again, I apologize for this inconvenience.

Comment · 3 years ago
It won't be any hard, as we simply gather players' positions and generate stats accordingly. The number of players or the size of the pitch doesn't affect the process.

I'm interested in trying a hockey video. You can contact me here if you are interested as well: [email protected]

Comment · 3 years ago
Thanks! Well flask, streamlit, plotly, opencv, docker, a trained object detection model (Faster RCNN or YOLO, there is not much difference). Google Cloud Run, some storage services on GCP, full python, react for front-end, and that's all.
Comment · 3 years ago
It doesn't work real time yet, the analysis is made after the video is completed. But it'd be a good feature to add.

Didn't considered gambling, how it would be? It'd be ike generating win or goals probabilities per minute?

Comment · 3 years ago
Thank you so much for your insightful comment. I genuinely appreciate the time you took to share your knowledge and experiences.

I'll take your advice about keeping this project as a passion rather than being desperate for profits seriously. Certainly, I'll need to face several issues, such as budget constraints, data, and process issues, like your anecdote about Yaya Toure's nearly supersonic shot.

All you mentioned will help a lot to make it thrive, especially the opportunities you found in high school football.

I'm grateful for the guidance you've provided to me. Thanks again!

Comment · 3 years ago
Comment · 3 years ago
Great, thanks!

To have access just fill out this form: * https://forms.gle/U8UeeTwrWiMjiUzZA

I'll send you some requirements for the video. I'm providing soccer video analysis at no cost in exchange for feedback.

Comment · 3 years ago
Comment · 3 years ago
Comment · 3 years ago
Thanks!

And the football commentator is amazing! To be honest, I wasn't aware of this. Exciting advancements in the soccer industry are being made possible by AI today.

I'm eager to give it a try and show it off to my friends!

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton