R/vision_utils.R

In torchvision: Models, Datasets and Transformations for Images

#' @importFrom magrittr %>%
NULL

#' A simplified version of torchvision.utils.make_grid
#'
#' Arranges a batch of (image) tensors in a grid, with optional padding between
#'   images. Expects a 4d mini-batch tensor of shape (B x C x H x W).
#'
#' @param tensor tensor to arrange in grid.
#' @param scale whether to normalize (min-max-scale) the input tensor.
#' @param num_rows number of rows making up the grid (default 8).
#' @param padding amount of padding between batch images (default 2).
#' @param pad_value pixel value to use for padding.
#'
#' @export
vision_make_grid <- function(tensor,
           scale = TRUE,
           num_rows = 8,
           padding = 2,
           pad_value = 0) {

    min_max_scale <- function(x) {
      min = x$min()$item()
      max = x$max()$item()
      x$clamp_(min = min, max = max)
      x$add_(-min)$div_(max - min + 1e-5)
      x
    }
    if(scale) tensor <- min_max_scale(tensor)

    nmaps <- tensor$size(1)
    xmaps <- min(num_rows, nmaps)
    ymaps <- ceiling(nmaps / xmaps)
    height <- floor(tensor$size(3) + padding)
    width <- floor(tensor$size(4) + padding)
    num_channels <- tensor$size(2)
    grid <-
      tensor$new_full(c(num_channels, height * ymaps + padding, width * xmaps + padding),
                      pad_value)
    k <- 0

    for (y in 0:(ymaps - 1)) {
      for (x in 0:(xmaps - 1)) {
        if (k >= nmaps)
          break
        grid$narrow(
          dim = 2,
          start =  1 + torch::torch_tensor(y * height + padding, dtype = torch::torch_int64())$sum(dim = 1),
          length = height - padding
        )$narrow(
          dim = 3,
          start = 1 + torch::torch_tensor(x * width + padding, dtype = torch::torch_int64())$sum(dim = 1),
          length = width - padding
        )$copy_(tensor[k + 1, , ,])
        k <- k + 1
      }
    }

    grid
}


#' Draws bounding boxes on image.
#'
#' Draws bounding boxes on top of one image tensor
#'
#' @param image : Tensor of shape (C x H x W) and dtype uint8.
#' @param boxes : Tensor of size (N, 4) containing bounding boxes in (xmin, ymin, xmax, ymax) format. Note that
#'            the boxes are absolute coordinates with respect to the image. In other words: `0  = xmin < xmax < W` and
#'            `0  = ymin < ymax < H`.
#' @param labels : character vector containing the labels of bounding boxes.
#' @param colors : character vector containing the colors
#'            of the boxes or single color for all boxes. The color can be represented as
#'            strings e.g. "red" or "#FF00FF". By default, viridis colors are generated for boxes.
#' @param fill : If `TRUE` fills the bounding box with specified color.
#' @param width : Width of text shift to the bounding box.
#' @param font : NULL for the current font family, or a character vector of length 2 for Hershey vector fonts.
# ' The first element of the vector selects a typeface and the second element selects a style.
#' @param font_size : The requested font size in points.
#'
#' @return  torch_tensor of size (C, H, W) of dtype uint8: Image Tensor with bounding boxes plotted.
#'
#' @examples
#' if (torch::torch_is_installed()) {
#' image <- torch::torch_randint(170, 250, size = c(3, 360, 360))$to(torch::torch_uint8())
#' x <- torch::torch_randint(low = 1, high = 160, size = c(12,1))
#' y <- torch::torch_randint(low = 1, high = 260, size = c(12,1))
#' boxes <- torch::torch_cat(c(x, y, x + 20, y +  10), dim = 2)
#' bboxed <- draw_bounding_boxes(image, boxes, colors = "black", fill = TRUE)
#' tensor_image_browse(bboxed)
#' }
#' @export
draw_bounding_boxes <- function(image,
                               boxes,
                               labels = NULL,
                               colors = NULL,
                               fill = FALSE,
                               width = 1,
                               font = c("serif", "plain"),
                               font_size = 10) {
  rlang::check_installed("magick")

  stopifnot("Image is expected to be a torch_tensor" = inherits(image, "torch_tensor"))
  stopifnot("Image is expected to be of dtype torch_uint8" = image$dtype == torch::torch_uint8())
  stopifnot("Pass individual images, not batches" = image$ndim == 3)
  stopifnot("Only grayscale and RGB images are supported" = image$size(1) %in% c(1, 3))
  stopifnot(
    "Boxes need to be in (xmin, ymin, xmax, ymax) format. Use torchvision$ops$box_convert to convert them" = (boxes[, 1] < boxes[, 3])$all() %>% as.logical() &&
      (boxes[, 2] < boxes[, 4])$all() %>% as.logical()
  )
  num_boxes <- boxes$shape[1]
  if (num_boxes == 0) {
    rlang::warn("boxes doesn't contain any box. No box was drawn")
    return(image)
  }
  if (!is.null(labels) && (num_boxes %% length(labels) != 0)) {
    rlang::abort(
      paste0(
        "Number of labels ",
        length(labels),
        "cannot be broadcasted on number of boxes",
        num_boxes
      )
    )
  }
  if (is.null(colors)) {
    colors <- grDevices::hcl.colors(n = num_boxes)
  }
  stopifnot("colors vector cannot be broadcasted on boxes" = num_boxes %% length(colors) == 0)

  if (!fill) {
    fill_col <- NA
  } else {
    fill_col <- colors
  }

  if (is.null(font)) {
    vfont <- c("serif", "plain")
  } else {
    if (is.null(font_size)) {
      font_size <- 10
    }
  }
  # Handle Grayscale images
  if (image$size(1) == 1) {
    image <- image$tile(c(4, 2, 2))
  }

  img_bb <- boxes$to(torch::torch_int64()) %>% as.array
  img_to_draw <- image$permute(c(2, 3, 1))$to(device = "cpu", dtype = torch::torch_long()) %>% as.array


  draw <- png::writePNG(img_to_draw / 255) %>%
    magick::image_read() %>%
    magick::image_draw()
  graphics::rect(img_bb[, 1],
       img_bb[, 2],
       img_bb[, 3],
       img_bb[, 4],
       col = fill_col,
       border = colors)
  if (!is.null(labels)) {
    graphics::text(
      img_bb[, 1] + width,
      img_bb[, 2] + width,
      labels = labels,
      col = colors,
      vfont = font,
      cex = font_size / 10
    )
  }
  grDevices::dev.off()
  draw_tt <-
    draw %>% magick::image_data(channels = "rgb") %>% as.integer %>% torch::torch_tensor(dtype = torch::torch_uint8())
  return(draw_tt$permute(c(3, 1, 2)))
}

#' Draw segmentation masks
#'
#' Draw segmentation masks with their respective colors on top of a given RGB tensor image
#'
#' @param image : torch_tensor of shape (3, H, W) and dtype uint8.
#' @param masks : torch_tensor of shape (num_masks, H, W) or (H, W) and dtype bool.
#' @param alpha : number between 0 and 1 denoting the transparency of the masks.
#   0 means full transparency, 1 means no transparency.
#' @param colors : character vector containing the colors
#'            of the boxes or single color for all boxes. The color can be represented as
#'            strings e.g. "red" or "#FF00FF". By default, viridis colors are generated for masks
#'
#' @return torch_tensor of shape (3, H, W) and dtype uint8 of the image with segmentation masks drawn on top.
#'
#' @examples
#' if (torch::torch_is_installed()) {
#' image <- torch::torch_randint(170, 250, size = c(3, 360, 360))$to(torch::torch_uint8())
#' mask <- torch::torch_tril(torch::torch_ones(c(360, 360)))$to(torch::torch_bool())
#' masked_image <- draw_segmentation_masks(image, mask, alpha = 0.2)
#' tensor_image_browse(masked_image)
#' }
#' @export
draw_segmentation_masks  <-  function(image,
                                      masks,
                                      alpha = 0.8,
                                      colors = NULL) {
  rlang::check_installed("magick")
  stopifnot("`image` is expected to be a torch_tensor" = inherits(image, "torch_tensor"))
  stopifnot("`image` is expected to be of dtype torch_uint8" = image$dtype == torch::torch_uint8())
  stopifnot("Pass individual images, not batches" = image$ndim == 3)
  stopifnot("Only grayscale and RGB images are supported" = image$size(1) %in% c(1, 3))
  if (masks$ndim == 2) {
    masks <- masks$unsqueeze(1)
  }
  stopifnot("`masks` must be of shape (H, W) or (num_masks, H, W)" = masks$ndim == 3)
  stopifnot("`masks` is expected to be of dtype torch_bool" = masks$dtype == torch::torch_bool())
  stopifnot("`masks` and `image` must have the same height and width" = masks$shape[2:3] == image$shape[2:3])
  num_masks <- masks$size(1)
  if (num_masks == 0) {
    rlang::warn("masks doesn't contain any mask. No mask was drawn")
    return(image)
  }
  if (is.null(colors)) {
    colors <- grDevices::hcl.colors(n = num_masks)
  }
  stopifnot("colors vector cannot be broadcasted on masks" = num_masks %% length(colors) == 0)

  out_dtype <- torch::torch_uint8()

  color_tt <-
    colors %>% grDevices::col2rgb() %>% t() %>% torch::torch_tensor(dtype = out_dtype)

  img_to_draw <- image$detach()$clone()

  colored_mask_stack <- torch::torch_stack(lapply(
     seq(masks$size(1)),
     function(x) color_tt[x, ]$unsqueeze(2)$unsqueeze(2)$mul(masks[x:x, , ])
     ),
    dim = 1
  )
  out <- img_to_draw * (1 - alpha) + torch::torch_sum(colored_mask_stack, dim = 1) * alpha
  return(out$to(out_dtype))
}


#' Draws Keypoints
#'
#' Draws Keypoints, an object describing a body part (like rightArm or leftShoulder), on given RGB tensor image.
#' @param image : Tensor of shape (3, H, W) and dtype uint8
#' @param keypoints : Tensor of shape (N, K, 2) the K keypoints location for each of the N detected poses instance,
#         in the format c(x, y).
#' @param connectivity : Vector of pair of keypoints to be connected (currently unavailable)
#' @param colors : character vector containing the colors
#'            of the boxes or single color for all boxes. The color can be represented as
#'            strings e.g. "red" or "#FF00FF". By default, viridis colors are generated for keypoints
#' @param radius : radius of the plotted keypoint.
#' @param width : width of line connecting keypoints.
#'
#' @return Image Tensor of dtype uint8 with keypoints drawn.
#'
#' @examples
#' if (torch::torch_is_installed()) {
#' image <- torch::torch_randint(190, 255, size = c(3, 360, 360))$to(torch::torch_uint8())
#' keypoints <- torch::torch_randint(low = 60, high = 300, size = c(4, 5, 2))
#' keypoint_image <- draw_keypoints(image, keypoints)
#' tensor_image_browse(keypoint_image)
#' }
#' @export
draw_keypoints <- function(image,
    keypoints,
    connectivity = NULL,
    colors = NULL,
    radius = 2,
    width = 3) {

  rlang::check_installed("magick")
  stopifnot("`image` is expected to be a torch_tensor" = inherits(image, "torch_tensor"))
  stopifnot("`image` is expected to be of dtype torch_uint8" = image$dtype == torch::torch_uint8())
  stopifnot("Pass individual images, not batches" = image$ndim == 3)
  stopifnot("Only RGB images are supported" = image$size(1) == 3)
  stopifnot("keypoints must be of shape (num_instances, K, 2)" = keypoints$ndim == 3)

  img_kpts <- keypoints$to(torch::torch_int64()) %>% as.array
  img_to_draw <- image$permute(c(2, 3, 1))$to(device = "cpu", dtype = torch::torch_long()) %>% as.array
  draw <- png::writePNG(img_to_draw / 255) %>%
    magick::image_read() %>%
    magick::image_draw()

  for (pose in dim(img_kpts)[[1]]) {
    graphics::points(img_kpts[pose,,1],img_kpts[pose,,2], pch = ".", col = colors, cex = radius)

  }
  # TODO need R-isation and vectorisation
    # for (kpt_id, kpt_inst in enumerate(img_kpts)) {
    #     if (connectivity) {
    #         for (connection in connectivity) {
    #             start_pt_x <- kpt_inst[connection[0]][0]
    #             start_pt_y <- kpt_inst[connection[0]][1]
    #
    #             end_pt_x <- kpt_inst[connection[1]][0]
    #             end_pt_y <- kpt_inst[connection[1]][1]
    #
    #             draw$line(
    #                 ((start_pt_x, start_pt_y), (end_pt_x, end_pt_y)),
    #                 widt = width,
    #             )
    #         }
    #     }
    # }
  grDevices::dev.off()
  draw_tt <-
    draw %>% magick::image_data(channels = "rgb") %>% as.integer %>% torch::torch_tensor(dtype = torch::torch_uint8())

    return(draw_tt$permute(c(3, 1, 2)))
}


#' Display image tensor
#'
#' Display image tensor onto the X11 device
#' @param image `torch_tensor()` of shape (1, W, H) for grayscale image or (3, W, H) for color image,
#'   of type `torch_uint8()` to display
#' @param animate support animations in the X11 display
#'
#' @export
tensor_image_display <- function(image, animate = TRUE) {
  stopifnot("`image` is expected to be of dtype torch_uint8" = image$dtype == torch::torch_uint8())
  stopifnot("Pass individual images, not batches" = image$ndim == 3)
  stopifnot("Only grayscale and RGB images are supported" = image$size(1) %in% c(1, 3))

  img_to_draw <- image$permute(c(2, 3, 1))$to(device = "cpu", dtype = torch::torch_long()) %>% as.array

  png::writePNG(img_to_draw / 255) %>% magick::image_read() %>% magick::image_display(animate = animate)

  invisible(NULL)
}


#' Display image tensor
#'
#' Display image tensor into browser
#' @param image `torch_tensor()` of shape (1, W, H) for grayscale image or (3, W, H) for color image,
#'   of type `torch_uint8()` to display
#' @param browser argument passed to [browseURL]
#'
#' @export
tensor_image_browse <- function(image, browser = getOption("browser")) {
  stopifnot("`image` is expected to be of dtype torch_uint8" = image$dtype == torch::torch_uint8())
  stopifnot("Pass individual images, not batches" = image$ndim == 3)
  stopifnot("Only grayscale and RGB images are supported" = image$size(1) %in% c(1, 3))

  img_to_draw <- image$permute(c(2, 3, 1))$to(device = "cpu", dtype = torch::torch_long()) %>% as.array

  png::writePNG(img_to_draw / 255) %>% magick::image_read() %>% magick::image_browse(browser = browser)

  invisible(NULL)
}