![]() ![]() ![]() Īfter step the structure becomes activity_recognition UCF-101. This will be our working directory from hereīefore step 2 the structure was activity_recognition UCF-101 train Archery v_Archery_g01_c01.avi v_Archery_g01_c02.avi. we will create a new directory called ‘activity_data’ where all the videos will be saved as frames. And video data set are usually saved in this format. this will help us to load the data frame by frame. The next step is to convert the video frames and save it as image frames.I will divide the dataset into train and test by approximately keeping 80% of the videos in each category in train section and remaining 20% in test section.Let’s prepare the dataset for making a clean data generator for this dataset I will be using 3 categories - Archery, Basketball and Biking from UCF-101 dataset. To elucidate this, I am going to consider the problem of activity recognition form videos using 3D CNN which takes a volume data (beside the spatial 2D data of image, the third dimension is the number of frames). Let’s consider the second scenario first, we need to train a model which takes a sequence input. In this post we will write data generator for second and third scenario. In the previous post we discussed how to write data generator for the first scenario and if you haven’t been through previous post, I would suggest that you go through that first - Chapter-2: Writing a generator function to read your data that can be fed for training an image classifier in Keras. Scenario-3: When you have multiple inputs that need to be fed to ensemble of model ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |