Page 199 - Ai_V1.0_Class9
P. 199

However, there are a few indicators you can look for:
                    • Artificial intelligence shows inconsistencies if observed closely, although it tries to piece together its creations
                   from  the  original  work.  The  artefacts  can  include  unnatural  blurriness,  inconsistent  lighting/shadowing,  or
                   repeating patterns, specially in the backgrounds.
                    • AI-generated images may include elements that seem unrealistic or improbable, such as impossible perspectives,
                   mismatched colours, or objects that defy physics, making the image appear unnatural or inconsistent with the
                   scene.
                    • Odd outlines to sharpen or smoothen the edges, stray pixels to cover inconsistency, and abnormal shapes can
                   be easily seen, if an image is zoomed to the maximum, on each of its parts.
                 Let's look at the concepts behind the generation of these images.


                         Supervised Learning and Discriminative Modelling


                 Supervised Learning is a type of Machine Learning where we teach models using examples that have labels. It
                 uses labelled datasets to train algorithms to predict outcomes and recognise patterns.
                 These labels tell the model the correct answer for each example. Discriminative Modelling is a special kind of
                 supervised learning that focuses on learning how to distinguish different classes. It looks at the features of the
                 data to figure out which class it belongs to.


                 Supervised Learning
                 Supervised learning is a machine learning where a model is trained on a labelled dataset, implying that each
                 input data point is associated with a corresponding output label. The goal of supervised learning is to learn the
                 mapping between input data and output labels, enabling the model to make predictions on new, unseen data.

                 Supervised learning is when we train the machine using labelled data. The machine is provided with a new set
                 of labelled data so that the supervised learning algorithm analyses the training data and generates the most
                 suitable and related outcome from the trained-labelled data. Labeled data contains data with the correct output
                 or classification. In simple words, input data is paired with the desired output thus making the machine learn to
                 predict the output for new input data.

                 For example, in the given images, first is the input image and characteristics of this image are marked as boy
                 and ball, which can be seen in center image. Now, according to supervised learning it has to learn the mapping
                 between input labels and output labels, which is shown in last image and highlights "ball" as red, "boy" as purple
                 and "boy playing with a ball" in a rectangle.

                                                                                                    Output
                               Input                     Features of given image
                                                                                               Label for the item
















                                                             Boy      Ball                   Boy playing with a ball


                                                                                  Introduction to Generative AI  197
   194   195   196   197   198   199   200   201   202   203   204