{"id":1451,"date":"2022-02-23T20:34:21","date_gmt":"2022-02-23T20:34:21","guid":{"rendered":"https:\/\/nilg.ai\/?p=1451"},"modified":"2025-03-17T12:39:39","modified_gmt":"2025-03-17T12:39:39","slug":"teaching-models-with-free-data","status":"publish","type":"post","link":"https:\/\/nilg.ai\/pt\/202202\/teaching-models-with-free-data\/","title":{"rendered":"Teaching Models With Free Data"},"content":{"rendered":"

\"\"<\/p>\n

\u201cThe more I see, the less I know\u201d might be a saying, but it does not apply to AI models. It\u2019s well known that the performance of an artificial neural network is highly dependent on the volume and on the diversity of the data that was shown to the model. This happens because exposing the models to diversity helps them select relevant features and mitigate potential bias, i.e. to understand the objects of study and to better perform their tasks.<\/p>\n

Getting data to feed such models might seem trivial since data is all around us. However, getting access to well-structured, labeled, copyright-free, and non-private data is still a pain for most data scientists.<\/p>\n

To overcome this, AI experts have come up with several approaches to optimize the learning process, in particular, smart architectures for Self Supervised Learning models.<\/p>\n

In this blog post, we (re)introduce you to Self Supervised Learning, along with one of our favorite strategies –\u00a0 Siamese Models <\/b>– and possible applications. If you want to learn even more about Self Supervised Learning and other Machine Learning techniques enroll to our online course:<\/p>\n

\n

 <\/p>\n

<\/h3>\n

What\u2019s Self Supervised Learning?<\/h3>\n

Supervised Learning is a machine learning approach that receives input data along with a specific target and aims to learn the data patterns and the transformation function that converts the input into the output. On the other side of the coin, we have Unsupervised Learning which is a method that needs no target to fulfill its mission, since it aims primarily to find patterns in data distributions.<\/p>\n

Then, we have Self Supervised Learning, which is an Unsupervised Learning method since it uses unlabeled data and has the particularity of creating synthetic labels to behave as a Supervised Learning model.<\/p>\n

These labels can be created by applying trivial transformations to the data. Here are some examples:<\/p>\n\n\n\n\n\n\n
\n

Objective<\/b><\/p>\n<\/td>\n

\n

Transformation<\/b><\/p>\n<\/td>\n

\n

Synthetic Label<\/b><\/p>\n<\/td>\n<\/tr>\n

\n

Quantify image rotation<\/p>\n<\/td>\n

\n

Random rotation<\/p>\n<\/td>\n

\n

Rotation angle<\/p>\n<\/td>\n<\/tr>\n

\n

Assess data Quality<\/p>\n<\/td>\n

\n

Insertion of random values into a data frame<\/p>\n<\/td>\n

\n

Binary flag for corrupted data<\/p>\n<\/td>\n<\/tr>\n

\n

Image Colorization<\/p>\n<\/td>\n

\n

RGB to black and white<\/p>\n<\/td>\n

\n

RGB image<\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

 <\/p>\n

The Self Supervised models can be used in two different manners:<\/p>\n