Page 334 - AI Ver 3.0 class 10_Flipbook
P. 334

• No coding required: You don’t need to write any code or understand Machine Learning Algorithms. Lobe’s
                interface is simple and visual, guiding you through the process step by step.


                       Introduction to Teachable Machine


              Teachable Machine is a powerful and user-friendly tool developed by Google in 2017. It is designed to make
              Artificial  Intelligence  (AI),  Machine  Learning  (ML),  and  Deep  Learning  (DL)  accessible  to  everyone,  including
              beginners with no technical background.
              It is built on top of TensorFlow.js, an open-source JavaScript library also developed by Google. TensorFlow.js
              allows Machine Learning models to run directly in a web browser, making it convenient and eliminating the need
              for specialised hardware or software installations.
              The users can train models using different types of input data:
                 • Images: Upload images from your computer or capture them live using a webcam. For example, you can train
                a model to distinguish between various objects, such as fruits or animals.
                 • Audio: Train the model to recognise sounds by providing audio samples, such as claps, whistles, or spoken words.
                 • Poses: Use a webcam to capture human poses, gestures, or movements for applications like fitness tracking or
                interactive games.

























              Following are the steps to use teachable machine:
               Step 1     Input Data Collection: Users start by providing training data. For instance:

                             • Upload pictures or use the webcam to capture images.
                             • Record audio clips or use existing sound files.
                             • Use the webcam to capture poses or gestures.
               Step 2     Model Training: Once the input data is ready, Teachable Machine uses Machine Learning algorithms

                          to identify patterns in the data and create a model. This process is entirely automated, so no coding
                          or technical expertise is required.
               Step 3     Model Testing: After training, the model can be tested in real time to check its performance. For

                          example:
                             • Point the webcam at an object to see if the model identifies it correctly.
                             • Make a gesture or pose and observe how the model reacts.



                    332     Touchpad Artificial Intelligence (Ver. 3.0)-X
   329   330   331   332   333   334   335   336   337   338   339