CS 7643 Quiz 3 Questions and Answers| Verified 100% Correct

CS 7643 Quiz 3 Questions and Answers| Verified 100% Correct Equivariance - ANSWER -If the input changes, the output changes in the same way if f(g(x) =g(f(x). If the beak of a bird in a picture moves a bit, the output values will move in the same way change to input causes equal change to output Invariance - ANSWER -If the input changes, the output stays the same. That is f(g(x)) = f(x) E.g. rotating/scaling a number will still result in that number being classified the same.. change to input does not affect output Useful if we care more about if a feature is present than exactly where it is saliency maps - ANSWER -Instead of using deconvnets, can, instead of taking the error gradient wrt model parameters, take gradient of class score we're trying to visualize wrt to the image itself (the input of the network) gets the degree to which a pixel contributed to that class score take absolute value of score because we care about degree, not direction helps us understand why model gives response they did Another method to make saliency maps: guided backpropagation algorithm (combination of deconvnet and gradient of class score wrt input of network) Sensitivity of loss to individual pixel changes , uses pre-softmax scores (gradient, then absval, then sum across channels) CAM = Class Activation Mapping - ANSWER -use Global Average Pooling layer as final layer to average the activations of each feature map and run through softmax loss layer to highlight the important regions of the image by projecting back the weights of the output on the convolutional feature maps Grad-CAM - ANSWER -more versatile version of CAM that can produce visual explanations for any arbitrary CNN, even if the network contains a stack of fully connected layers too let the gradients of any target concept score flow into the final convolutional layer; then compute an importance score based on the gradients and produce a coarse localization map highlighting the important regions in the image for predicting that concept What regions of image is model looking at to make prediction? Which individual regions have highest class activation as you extract layer from CNN? Direction/magnitude of gradients to determine which gradients are causing the most updates to the NN Objective: inspective given layer of CNN and correlate to output Task specific (if asked what is a dog - dog pixels are more important) Adversarial examples - ANSWER -Inputs formed by applying small but intentionally worst-case perturbations to examples from the dataset, such that the perturbed input results in the model outputting an incorrect ANSWER with high confidence. Guided Backprop - ANSWER -Layer by layer (deconvolution is similar to backprop) From details to more abstracted representations L1 Loss - ANSWER -Sum of Absolute Value of (true - predicted) L2 Loss - ANSWER -Sum of Absolute Value of (true - predicted)^2 Mean Squared Error (MSE) - ANSWER -Average of (true - predicted)^2 Style Transfer - ANSWER -measure the difference in style between the synthesized image and the style image Sum of Gram matrix of each layer Stype vs prediction Where G is the Gram matrix that abstracts the correlation between the layers Content Loss - ANSWER -the difference in content features between the synthesized image and the content image via the squared loss function Binary Cross-Entropy Loss - ANSWER --1* log(prediction of true class) Balanced Cross-Entropy Loss - ANSWER --1 * alpha * log(prediction of true class) Focal Loss - ANSWER --1 * (1- prediction of true class)^gamma * log(prediction of true class) Class Balanced Focal Loss - ANSWER --1 * alpha_t (1- prediction of true class)^gamma * log(prediction of true class) SSD (Single-Shot Detector) - ANSWER -grid as anchors w different scales/aspect ratios Based on VGG model till conv5_3 layer YOLO (You Only Look Once) - ANSWER -single-scale faster for same size Customized architecture, full connected at the end NMS before results R-CNN - ANSWER -Find regions of interest (ROIs) with object-like things. Slow -based on Selective Search; returns scores and bounding boxes; 100's of images/crops to process. Wasting a lot of computing resources for the same image portions. Classifier those regions and refine their bounding boxes Fast R-CNN - ANSWER -map each ROI in image to corresponding region in feature maps Reuse comp by finding regions in feature maps Feature extraction once per image Issue: variable input size to FC layers Solved with ROI Pooling Faster R-CNN - ANSWER -DL to do everything RPN (region proposal network) ... NN generates proposals select TopK of them outputs objectness score & bounding box Losses: 4 losses: bounding box loss, objectiveness score loss, regression loss, classifier loss for each class . anchors as grid Mask R-CNN - ANSWER -same as Faster R-CNN, but learns a mask that says which pixels touch the object