given rois, overlaps, gt labels, compute bounding box regression targets :param rois: roidb[i]['boxes'] k * 4 :param overlaps: roidb[i]['

given ground truth, prepare roidb :param box_list: [image_index] ndarray of [box_index][x1, x2, y1, y2] :param gt_roidb: [ima

inject deep into distutils to customize how the dispatch to gcc/nvcc works. If you subclass UnixCCompiler, it's not trivial to get your subcl

evaluate detection proposal recall metrics record max overlap value for each gt box; return vector of overlap values :param r

Find a file in a search path

Generate anchor (reference) windows by enumerating aspect ratios X scales wrt a reference (0, 0, 15, 15) window.

Generate detections results using RPN. :param predictor: Predictor :param test_data: data iterator, must be non-shuffled :param imdb:

filtering parameters. default skips that filter. :param catNms (str array) : get cats for given cat names :param supNms (str

return a dict of testbatch :param roidb: ['image', 'flipped'] + ['boxes'] :return: data, label, im_info

return a dict of testbatch :param roidb: ['image', 'flipped'] + ['boxes'] :return: data, label, im_info

return a dict of testbatch :param roidb: ['image', 'flipped'] :return: data, label, im_info

calculate label shape according to data shape :param data_shape: dict of shape {"data": (1, 3, 600, 1000)} :param fea

valid images have at least 1 fg or bg roi

valid images have at least 1 fg or bg roi

Load cats with the specified ids. :param ids (int array) : integer ids specifying cats :return: cats (object array) : l

Load anns with the specified ids. :param ids (int array) : integer ids specifying img :return: imgs (object array) : lo

Convert result data from a numpy array [Nx7] where each row contains {imageID,x1,y1,w,h,score,class} :param data (numpy.ndarray)

for a given index, load image and bounding boxes info from a single image list :return: list of record['boxes', 'gt_classes', 'ins_id

load ground truth roidb

find out which indexes correspond to given image set (train or val) :return:

load proposal roidb (append_gt when training)

Locate the CUDA environment on the system Returns a dict with keys 'home', 'nvcc', 'include', and 'lib64' and values giving the absolute path

wrapper for calculating offline validation for faster data analysis in this example, all threshold are set by hand :param predictor: Pred

wrapper for calculating offline validation for faster data analysis in this example, all threshold are set by hand :param predictor: Pred

only resize input image to target size and return scale :param im: BGR image input by opencv :param target_size: one dimensional size (th

generate random sample of ROIs comprising foreground and background examples :param rois: all_rois [n, 4]; e2e: [n, 5] with batch_index :

Checkpoint the model data into file. :param prefix: Prefix of model name. :param epoch: The epoch number of the model. :param arg_params:

Compute and display summary metrics for evaluation results. Note this functin can *only* be applied on the default parameter setting

transform into mxnet tensor substract pixel size and transform to correct format :param im: [height, width, channel] in BGR :param pi

visualize all detections in one image :param im_array: [b=1 c h w] in rgb :param detections: [ numpy.ndarray([[x1 y1 x2 y2 score]]) for j

decalre dependency and inplace optimization options