Alex Bodner's banner

Alex Bodner

@AlexBodner_ • 1,516 subscribers

open source @roboflow | AI engineering at @UdeSA🇦🇷 prev founder @SatsOnFire

Shorts

I timed a 100m race from the tv broadcast. No timing hardware, no finish sensors, nothing trained. All open source vision libraries. we called the winner at 11.26s. the official time was 11.23. the stack: - trackers (BoT-SORT + camera motion compensation) to hold all 8 runners through the pan - RF-DETR for the runner detections - opencv to find the finish line, because it's just white paint on the track - fft magic to hear the starting gun on the soundtrack tell me if you want the tutorial on how i built it!

I timed a 100m race from the tv broadcast. No timing hardware, no finish sensors, nothing trained. All open source vision libraries. we called the winner at 11.26s. the official time was 11.23. the stack: - trackers (BoT-SORT + camera motion compensation) to hold all 8 runners through the pan - RF-DETR for the runner detections - opencv to find the finish line, because it's just white paint on the track - fft magic to hear the starting gun on the soundtrack tell me if you want the tutorial on how i built it!

98,817 просмотров

as Satya pointed out, using Astra to track is expensive, and classic trackers already do the job. so i only called it when the track was lost 10 API calls for a 269 frame video. Next I will be trying to track literally anything out of a prompt. What are the most complex cases you think we can tackle now?

as Satya pointed out, using Astra to track is expensive, and classic trackers already do the job. so i only called it when the track was lost 10 API calls for a 269 frame video. Next I will be trying to track literally anything out of a prompt. What are the most complex cases you think we can tackle now?

19,149 просмотров

Anthropic published a way to read LLM’s inner thoughts last week (Natural Language Autoencoders). 2 days later, we spent 36h straight hours at Platanus Ventures Hack Buenos Aires, were we built an open-source system that uses it to detect deception and steer models back into alignment. Here’s what we found:

Anthropic published a way to read LLM’s inner thoughts last week (Natural Language Autoencoders). 2 days later, we spent 36h straight hours at Platanus Ventures Hack Buenos Aires, were we built an open-source system that uses it to detect deception and steer models back into alignment. Here’s what we found:

15,943 просмотров

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

Astra is Messi level🤖⚽️🐐 AI could solve Navier-Stokes, can it solve football? I froze the video before memorable plays and made Astra, Fable and Sol take key decisions seeing Messi's POV Compared it with what Messi did and Astra was the only one choosing the Messi options. What should I try next?

Astra is Messi level🤖⚽️🐐 AI could solve Navier-Stokes, can it solve football? I froze the video before memorable plays and made Astra, Fable and Sol take key decisions seeing Messi's POV Compared it with what Messi did and Astra was the only one choosing the Messi options. What should I try next?

17,711 просмотров • 4 дней назад

Giving Microduck object-tracking vision capabilities! 🦆👀🤖 It can now play fetch and doesn't get confused by other objects. Using Roboflow Trackers library I made Microduck keep track of the different objects in its sight and has a control policy for reaching the ball it tracks. Open-Source code below 👇

Giving Microduck object-tracking vision capabilities! 🦆👀🤖 It can now play fetch and doesn't get confused by other objects. Using Roboflow Trackers library I made Microduck keep track of the different objects in its sight and has a control policy for reaching the ball it tracks. Open-Source code below 👇

31,283 просмотров • 12 дней назад

and my favorite part: pass detection and suggestion. Used BoT-SORT tracker, ball and player detectors, and keypoint detection for the pitch. Tell me if you want the tutorial on how I built the full thing!

and my favorite part: pass detection and suggestion. Used BoT-SORT tracker, ball and player detectors, and keypoint detection for the pitch. Tell me if you want the tutorial on how I built the full thing!

35,923 просмотров • 23 дней назад

I’m thrilled to present the KAN Convolutional Layer, a promising neural architecture for image processing. Last week KANs came out, as an alternative to the MLP. We extended this idea to Convolutional Layers, creating the KAN Convolution. Join me in this thread to more about it🧵

I’m thrilled to present the KAN Convolutional Layer, a promising neural architecture for image processing. Last week KANs came out, as an alternative to the MLP. We extended this idea to Convolutional Layers, creating the KAN Convolution. Join me in this thread to more about it🧵

201,550 просмотров • 2 лет назад

🔥Convolutional Kolmogorov-Arnold Networks Preprint out now🔥 I'm very excited to release our preprint on Arxiv. Also, my first paper. We proposed a novel architecture that shows potential to compete with CNNs. Want to learn more about it? 🧵Follow this thread👇

🔥Convolutional Kolmogorov-Arnold Networks Preprint out now🔥 I'm very excited to release our preprint on Arxiv. Also, my first paper. We proposed a novel architecture that shows potential to compete with CNNs. Want to learn more about it? 🧵Follow this thread👇

102,833 просмотров • 2 лет назад

Больше нет контента для загрузки