brails.processors.nfloors_detector.lib.utils.utils module

class brails.processors.nfloors_detector.lib.utils.utils.CustomDataParallel(module, num_gpus)

Bases: DataParallel

force splitting data to all gpus instead of sending all data to cuda:0 and then moving around.

scatter(inputs, kwargs, device_ids)

brails.processors.nfloors_detector.lib.utils.utils.aspectaware_resize_padding(image, width, height, interpolation=None, means=None)

brails.processors.nfloors_detector.lib.utils.utils.display(preds, imgs, obj_list, imshow=True, imwrite=False)

brails.processors.nfloors_detector.lib.utils.utils.from_colorname_to_bgr(color)

brails.processors.nfloors_detector.lib.utils.utils.get_index_label(label, obj_list)

brails.processors.nfloors_detector.lib.utils.utils.get_last_weights(weights_path)

brails.processors.nfloors_detector.lib.utils.utils.init_weights(model)

brails.processors.nfloors_detector.lib.utils.utils.invert_affine(metas: float | list | tuple, preds)

brails.processors.nfloors_detector.lib.utils.utils.plot_one_box(img, coord, label=None, score=None, color=None, line_thickness=None)

brails.processors.nfloors_detector.lib.utils.utils.postprocess(x, anchors, regression, classification, regressBoxes, clipBoxes, threshold, iou_threshold)

brails.processors.nfloors_detector.lib.utils.utils.preprocess(*image_path, max_size=512, mean=(0.406, 0.456, 0.485), std=(0.225, 0.224, 0.229))

brails.processors.nfloors_detector.lib.utils.utils.preprocess_video(*frame_from_video, max_size=512, mean=(0.406, 0.456, 0.485), std=(0.225, 0.224, 0.229))

brails.processors.nfloors_detector.lib.utils.utils.replace_w_sync_bn(m)

brails.processors.nfloors_detector.lib.utils.utils.standard_to_bgr(list_color_name)

brails.processors.nfloors_detector.lib.utils.utils.variance_scaling_(tensor: Tensor, gain: float = 1.0) → Tensor

initializer for SeparableConv in Regressor/Classifier reference: https://keras.io/zh/initializers/ VarianceScaling