I need the answer as soon as possible

Question 3: (10 marks) You are given a CNN as a backbone/feature-extractor that takes a 120 x 120 RGB as input and outputs a (55, 34,34) feature map. You are required to add a convolutional layer on top of this feature map so that the output of your convolutional layer is (1000,1,1). As the answer to this question, give the following details: A. The number of filters in your proposed layer B. The number of channels in each filter C. The total number of trainable parameters in your layer, assuming that the filters are trainable. D. The activation function to use if you want to treat the output of your layer as a probability distribution over 1000 classes