AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Problem of Unsupervised Running in Machine Oning
Machine learning has been unbelievably successful in a natural language processing. That success has not yet translated in the domain of vision, although im predicting that it will happen very soon. A andander is to take analogy rite the fact that every sample in so of self supervised contect give you o more information than er a supervised sample. And therefore, why not use enabled it in large quantity to pre train very large normal nets so that we can find tun them for the task that we are interested in?