Dump relative paths. Args: - jpg_dir: - txt_output_dir: Returns: - txt_save_fpath:

Compare classnames in *_names.txt file against tsv column entries. For training datasets, these must be *exactly* the same.

Args: - args: - use_gpu Returns: - None

Return the index of this element in the list

flatten_list

Form a large dictionary mapping (parent,filename)->(update objects). Later we can check to see if this image is in our big dictionary of filename

Args: - n_classes: integer representing number of output classes - load_imagenet_model: whether to initialize from ImageN

Write dummy camvid data. Expect inter [4,2,1] Expect union [7,5,1] IoUs: 4/7, 2/5, 1/1

Find the classes to exclude when evaluating a "relabeled" MSeg model on the val split of a training dataset. We retrieve the dictionary `id_t

Returns: ndarray: the visualized image of shape (H, W, 3) (RGB) in uint8 type. The shape is scaled w.r.t the input

Args: - Returns: -

Args: - mask Returns: - coordinate of mean pixel location as (x,y)

Args: - label_img: Returns: - list of strings, representing classnames

Args: - split: string representing training, validation, or testing split of the data - fname_stem: Returns: - list of strings, repre

Uses the raw version of the dataset (as originally distributed) to get a mask. Args: - seq_id: sequence ID, parent of file name in file syst

Loads a state_dict created by an earlier call to state_dict(). If ``fp16_optimizer_instance`` was constructed from some ``init_optimi

Copy master parameters to model parameters. Args: model_params: List of model parameters created by :func:`prep_param_lists`.

new_like()

Swap the pixel values inside a label map's mask to a new value. This effectively changes the mask's category on disk. Get fname stem from rgb ima

Network input should be at least crop size, so we pad using mean values if provided image is too small. No rescaling is performed here. We use cv

# folder containing subfolders as 360/720/1080 Args: folder resolution scale: string representing inference scale option, either 'ss

Dump per-class IoUs and mIoU to stdout.

Args: rgb (bool): whether to return RGB colors or BGR colors. maximum (int): either 255 or 1 Returns: ndarray: a vec

Copy the data out of a csv file, as a list of OrderedDicts. Args: - csv_fpath: string representing path to a csv file. Rows: - rows: list o

Args: - old_fpath - new_fpath Returns: - None

Args: - old_fpath - new_fpath Returns: - None

Args: - fpath Returns: - rgb_img_fpaths - label_fpaths

No need to copy the RGB files again. We just update the label file paths. Args: - old_dataroot: - new_dataroot: - orig_pair: Tuple contain

Given path to a dataset, given names of _names.txt Remap according to the provided tsv. (also account for the fact that 255 is always unlabeled)

reset_hidden()

Args: - args: - use_gpu

Args: - img_rgb: Array representing 3-channel image in RGB order - label_img: Array representing grayscale image, where intensities correspond

First, pad the image. If input is (384x512), then we must pad it up to shape to have shorter side "scaled base_size". Then we perform the slidi

For visualizing classnames in an image. When we wish to place text over a mask, for nonconvex regions, we cannot use mask pixel mean location (ma

Given a list of work, and a desired number of n workers, launch n worker processes that will each process 1/nth of the total work. Args: - num

Returns a dict containing the current state of this :class:`FP16_Optimizer` instance. This dict contains attributes of :class:`FP16_O

Performs a single optimization step. Arguments: closure (callable, optional): A closure that reevaluates the model

Make sure 255 stays mapped to 255 at each level (to be ignored in cross-entropy loss).

Test a model that was trained in the exact same taxonomy we wish to evaluate in. Args: - args: - use_gpu

Args: - pred - target - classes Returns: - None

Ensure pixel values inside a label map's mask were previously swapped to a new value. Mask's category on disk should have previously been changed.

Vertically stack an image and a palette image, placing the palette image below it. Args: - top_img - palette_img Returns: - vstack_img

Create a 6-image tile grid with the following structure: ------------------------------------------------------------ RGB Image | Blended

Note that this function will add a carriage return, so please be mindful of this. Freshly loaded lines from a file will have carriage returns, b