Multibox Loss Args: predictions (tuple): A tuple containing loc preds, conf preds, and prior boxes from SSD net.

Decode locations from predictions using priors to undo the encoding we did for offset regression at train time. Args: loc (tensor): lo

Upsample and add two feature maps. Args: x: (Variable) top feature map to be upsampled. y: (Variable) lateral feature map.

Encode the variances from the priorbox layers into the ground truth boxes we have matched (based on jaccard overlap) with the prior boxes. Arg

Compute the jaccard overlap of two sets of boxes. The jaccard overlap is simply the intersection over union of two boxes. Here we operate on

Upsample and add two feature maps. Args: x: (Variable) top feature map to be upsampled. y: (Variable) lateral feature map.

Convert prior_boxes to (cx, cy, w, h) representation for comparison to center-size form ground truth data. Args: boxes: (tensor) poin

3x3 convolution with padding

Compute the jaccard overlap of two sets of boxes. The jaccard overlap is simply the intersection over union of two boxes. Args: box_a

Match each prior box with the ground truth box of the highest jaccard overlap, encode the bounding boxes, then return the matched indices corr

Convert prior_boxes to (xmin, ymin, xmax, ymax) representation for comparison to point form ground truth data. Args: boxes: (tensor)

S3FD: Single Shot Scale-invariant Face Detector

roi pool on pyramid feature maps

focusing is parameter that can adjust the rate at which easy examples are down-weighted. alpha may be set by inve

Constructs a ResNet-50 model. Args: pretrained (bool): If True, returns a model pre-trained on ImageNet

We resize both tensors to [A,B,2] without new malloc: [A,2] -> [A,1,2] -> [A,B,2] [B,2] -> [1,B,2] -> [A,B,2] Then we compute the area of

Utility function for computing log_sum_exp while determining This will be used to determine unaveraged confidence loss across all examples in

Apply non-maximum suppression at test time to avoid detecting too many overlapping bounding boxes for a given object. Args: boxes: (te

Returns the original annotation of image at index Note: not using self.__getitem__(), as any transformations passed in could mess up

Returns the original image object at index in PIL form Note: not using self.__getitem__(), as any transformations passed in could mes

Returns the original image object at index in PIL form Note: not using self.__getitem__(), as any transformations passed in could mes

Returns the original image object at index in PIL form Note: not using self.__getitem__(), as any transformations passed in could mes

Args: index (int): Index Returns: tuple: Tuple (image, target, height, width). target is t

Match each prior box with the ground truth box of the highest jaccard overlap, encode the bounding boxes, then return the matched indices corr

Constructs a ResNet-101 model. Args: pretrained (bool): If True, returns a model pre-trained on ImageNet

Draw detected bounding boxes.

Args: image (Tensor): image tensor to be transformed Return: a tensor with channels swapped according to swap

Arguments: target (annotation) : the target annotation to be made usable will be an ET.Element Returns:

Args: target (dict): COCO target json annotation as a python dict height (int): height width (int): width

Arguments: target (annotation) : the target annotation to be made usable will be an ET.Element Returns:

Args: index (int): Index Returns: tuple: Tuple (image, target). target is the object retur