Feat: add item_transform_perspective for detection and segmentation items

[DerrickUnleashed]

Added item_transform_perspective for detection item, dataset and segmentation item, dataset

Closes #353

[DerrickUnleashed]

# perspective corners covering the full image
startpoints <- function(img) {
  h <- img$shape[2]; w <- img$shape[3]
  list(c(0, 0), c(w - 1, 0), c(w - 1, h - 1), c(0, h - 1))  # TL, TR, BR, BL
}
# warp: pinch the top edge inward, push the bottom edge down/out
endpoints <- function(img) {
  h <- img$shape[2]; w <- img$shape[3]
  list(c(150, 60), c(w - 151, 40), c(w - 1, h - 1), c(0, h - 1))
}

url <- "https://upload.wikimedia.org/wikipedia/commons/b/b6/Felis_catus-cat_on_snow.jpg"

# ========== 1. DETECTION ITEM ==========
img <- base_loader(url) |> transform_to_tensor()
boxes <- torch_tensor(matrix(c(600, 200, 2880, 1860), ncol = 4), dtype = torch_float32())
det_item <- list(x = img, y = list(boxes = boxes, labels = "CAT"))
class(det_item) <- c("image_with_bounding_box", "list")

det_warped <- item_transform_perspective(det_item, startpoints(img), endpoints(img))

p1 <- draw_bounding_boxes(det_item, colors = "blue", width = 10)$to(torch_float())$div(255)
p2 <- draw_bounding_boxes(det_warped, colors = "red", width = 10)$to(torch_float())$div(255)
grid1 <- vision_make_grid(torch_stack(list(p1, p2)), scale = TRUE)
tensor_image_browse(grid1)

# ========== 2. DETECTION DATASET ==========
ds_det <- pascal_detection_dataset(year = "2007", split = "trainval",
                                   transform = transform_to_tensor, download = TRUE)
orig_det <- ds_det[1]
ds_det_warped <- item_transform_perspective(ds_det, startpoints(orig_det$x), endpoints(orig_det$x))
warp_det <- ds_det_warped[1]

p3 <- draw_bounding_boxes(orig_det, colors = "blue", width = 5)$to(torch_float())$div(255)
p4 <- draw_bounding_boxes(warp_det, colors = "red", width = 5)$to(torch_float())$div(255)
grid2 <- vision_make_grid(torch_stack(list(p3, p4)), scale = TRUE)
tensor_image_browse(grid2)

# ========== 3. SEGMENTATION ITEM ==========
img2 <- base_loader(url) |> transform_to_tensor()
h <- img2$shape[2]; w <- img2$shape[3]

mask1 <- torch_zeros(h, w, dtype = torch_bool())
mask1[, 1:150] <- TRUE

mask2 <- torch_zeros(h, w, dtype = torch_bool())
mask2[, 350:500] <- TRUE

masks <- torch_stack(list(mask1, mask2))
seg_item <- list(x = img2, y = list(masks = masks, labels = torch_tensor(c(1L, 2L)),
                                    image_height = h, image_width = w))
class(seg_item) <- c("image_with_segmentation_mask", "list")

seg_warped <- item_transform_perspective(seg_item, startpoints(img2), endpoints(img2))

p5 <- draw_segmentation_masks(seg_item, alpha = 0.5, colors = c("red", "blue"))$to(torch_float())$div(255)
p6 <- draw_segmentation_masks(seg_warped, alpha = 0.5, colors = c("red", "blue"))$to(torch_float())$div(255)
grid3 <- vision_make_grid(torch_stack(list(p5, p6)), scale = TRUE)
tensor_image_browse(grid3)

# ========== 4. SEGMENTATION DATASET ==========
ds_seg <- pascal_segmentation_dataset(year = "2007", split = "trainval",
                                      transform = transform_to_tensor, download = TRUE)
orig_seg <- ds_seg[1]
ds_seg_warped <- item_transform_perspective(ds_seg, startpoints(orig_seg$x), endpoints(orig_seg$x))
warp_seg <- ds_seg_warped[1]

p7 <- draw_segmentation_masks(orig_seg, alpha = 0.5)$to(torch_float())$div(255)
p8 <- draw_segmentation_masks(warp_seg, alpha = 0.5)$to(torch_float())$div(255)
grid4 <- vision_make_grid(torch_stack(list(p7, p8)), scale = TRUE)
tensor_image_browse(grid4)

DETECTION ITEM

DETECTION DATASET

SEGMENTATION ITEM

SEGMENTATION DATASET

[cregouby]

praise @DerrickUnleashed, this looks nice.
todo please resolve merge conflicts.

[DerrickUnleashed]

praise @DerrickUnleashed, this looks nice. todo please resolve merge conflicts.

Thanks @cregouby , I have resolved the conflicts