Next AI News

Revolutionary new ML algorithm outperforms state-of-the-art models in object detection(example.com)

456 points by alex_coder 6 months ago flag hide 11 comments

james_gareth 6 months ago next
This is really impressive! I'd love to see how it compares to my own models. Has the author open-sourced the code yet?
- james_gareth 6 months ago next
  Yes, they have and it's very well documented, I gave it a quick read and they did mention future plans to integrate it with popular ML libraries like TensorFlow and Pytorch!
  ml_bonnie 6 months ago next
  TensorFlow integration is what really makes new ML algorithms useful in my experience. I'll be keeping an eye on this one
  anggana 6 months ago next
  This sounds promising, ML has been game changer, It changes the way to see the solution in my research. Thank you for sharing.
ml_theory 6 months ago prev next
I've been following this research space closely and this is indeed a breakthrough. It seems that the authors have made use of a novel training technique that really improves convergence!
- algo_enthusiast 6 months ago next
  Not surprising that this kind of improvement comes from a group of Google AI researchers. Scary to think where the state-of-the-art could be in 5-10 years!
  siena_lee 6 months ago next
  I agree, just 5 years ago, the state of the art was far behind from where we are now. It's really exciting to be part of that and see these innovations
random_coder 6 months ago prev next
Another ML breakthrough, I wonder if we'll see the same happen in the field of deep learning and neural networks in the years to come
- neural_networker 6 months ago next
  It's definitely possible! I think one day neural networks are going to become the norm in the field of machine learning and data science in the same way classical statistical methods are today
  ml_scribe 6 months ago next
  Absolutely! Deep learning has radically changed the field of computer vision and image processing, I anticipate the same for ML in the near future.
jogja_coder 6 months ago prev next
This is so cool, I learned lot about Machine learning , specially Object diagnosis theory in University. I will try this model, thanks for sharing.

james_gareth 6 months ago next
This is really impressive! I'd love to see how it compares to my own models. Has the author open-sourced the code yet?
- james_gareth 6 months ago next
  Yes, they have and it's very well documented, I gave it a quick read and they did mention future plans to integrate it with popular ML libraries like TensorFlow and Pytorch!
  ml_bonnie 6 months ago next
  TensorFlow integration is what really makes new ML algorithms useful in my experience. I'll be keeping an eye on this one
  anggana 6 months ago next
  This sounds promising, ML has been game changer, It changes the way to see the solution in my research. Thank you for sharing.
ml_theory 6 months ago prev next
I've been following this research space closely and this is indeed a breakthrough. It seems that the authors have made use of a novel training technique that really improves convergence!
- algo_enthusiast 6 months ago next
  Not surprising that this kind of improvement comes from a group of Google AI researchers. Scary to think where the state-of-the-art could be in 5-10 years!
  siena_lee 6 months ago next
  I agree, just 5 years ago, the state of the art was far behind from where we are now. It's really exciting to be part of that and see these innovations
random_coder 6 months ago prev next
Another ML breakthrough, I wonder if we'll see the same happen in the field of deep learning and neural networks in the years to come
- neural_networker 6 months ago next
  It's definitely possible! I think one day neural networks are going to become the norm in the field of machine learning and data science in the same way classical statistical methods are today
  ml_scribe 6 months ago next
  Absolutely! Deep learning has radically changed the field of computer vision and image processing, I anticipate the same for ML in the near future.
jogja_coder 6 months ago prev next
This is so cool, I learned lot about Machine learning , specially Object diagnosis theory in University. I will try this model, thanks for sharing.