Today, #LIDAR is used in all autonomous cars except in Tesla

Lidar sensors are big, bulky, expensive, and ugly to look at. Not only that, they do a poor job in snow, sleet, hail, smoke, and smog. If you can’t see the road ahead, neither can LIDAR!.

That last part is one of the reasons Elon Musk refuses to incorporate lidar sensors into the self-driving hardware package for Tesla cars.

Apple & Cornell University have solved the problem of depth precision and this paves the way for faster adoption for safer yet cheaper cars!

Read more here: https://lnkd.in/dZgS6id
Research paper: https://lnkd.in/djRhzq3
#research #selfdriving #deeplearning

✴️ @AI_Python_EN

I am implementing a training loop that can be used with Auxiliary Classifier. So what is an Auxiliary Classifier?

Auxiliary Classifier are the ones in which we take the outputs of layers of some previous layers along with the final outputs and compare it with the targets and calculate a loss based on both the outputs from the final layer as well as the previous layer.

How does this help?
I think before even me saying how this is going to be helpful, I think this intuitively gives an idea of how is this going to aid the training process, I got so freaking excited when I came to know about this.

So, How does this help?
- Solves the gradient Vanishing problem
- Low-level features get more and more accurate and thus making the model more and more accurate.
- This also acts as regularization, it kind of can be thought as putting some constraints on the model which help in regularization.

I am not sure which paper first introduced Auxiliary Approaches, but I am trying to train an FCN using this, let's see how this aids the process :)
Maybe the Inception paper.

#machinelearning #deeplearning #datascience #python #artificialintelligence #selfdriving #nlp #computervision

✴️ @AI_Python_EN