torchvision 0.10.0

Breaking changes

• model_maskrcnn_*() now return image-sized mask probabilities in masks, shape (N, H, W) (@sebffischer, #412). Code that used the per-box (N, 28, 28) probabilities should read output$detections[[1]]$mask_probs instead of output$detections[[1]]$masks. For image overlays, use output$detections[[1]]$masks > 0.5.
• State dictionaries saved with torchvision 0.9.0 need migration for model_vit_*(), model_facenet_inception_resnet_v1(), model_convnext_*_upernet(), model_fasterrcnn_resnet50_fpn_v2() and model_maskrcnn_resnet50_fpn_v2(): parameters were added, removed or renamed, so these checkpoints cannot be loaded unchanged with strict = TRUE (#405, #406, #410, #412). RF-DETR checkpoints from earlier development versions may also need migration after changes to decoder depths and resolutions (#413). There is no automatic migration for user-saved checkpoints. Keep the package version that created them, or explicitly migrate and validate the weights. model_maxvit() retains its state-dict keys and shapes, but its corrected forward computation can change predictions (#405). Support for the hosted pretrained weights does not imply compatibility with user-saved checkpoints.
• When supplied, draw_bounding_boxes() labels and colors must have length 1 or match the number of original boxes. Partial recycling is no longer supported (#415).

New models

• Added RF-DETR model family: model_rfdetr_*, with * being nano(), small(), medium(), base(), base_2(), base_o365() and large() for real-time object detection with DINOv2 backbone and deformable-attention decoder (@DerrickUnleashed, #327).
• Added LW-DETR model family: model_lw_detr_*, with * being tiny(), small(), medium() and large() for real-time object detection with ViT backbone and deformable-attention decoder (@srishtiii28, #328).

New datasets

• Added cityscapes_dataset() relying on manual download due to usage license (#279, @ANAMASGARD).

New features

• Added target_transform_resize to manage bounding_box resizing in parallel to transform_resize for images (#337).
• Added target_transform_rotate() for converting bounding boxes from xyxy to xyxyr format with a rotation angle (@DerrickUnleashed, #338). Both operate on detection item targets (the y part of a dataset item) for composable target transform pipelines.
• Added a new draw_bounding_boxes S3 method for image_with_rotated_box that renders rotated boxes as polygons (@DerrickUnleashed, #338).
• Added item_transform_hflip() for horizontally flipping dataset items, with support for detection and segmentation item types and datasets (@DerrickUnleashed, #366).
• Added item_transform_vflip() for vertically flipping dataset items, with support for detection and segmentation item types and datasets (@DerrickUnleashed, #369).
• Added SAHI (Slicing Aided Hyper Inference) support via a three-function pipeline: prepare_sahi_split() precomputes overlapping crop windows, then transform_sahi_crop() slices images and target_transform_sahi_crop() adjusts detection targets per crop. (@DerrickUnleashed, #324)
• Added item_transform_affine() and target_transform_affine() for applying an affine transformation to a dataset item and its detection target, with support for detection and segmentation items and datasets (#367).
• Added item_transform_center_crop() for cropping images from the center to a specified size for dataset items, with support for detection and segmentation item types and datasets (@DerrickUnleashed, #370).
• Added item_transform_random_horizontal_flip() and item_transform_random_vertical_flip() for random geometric augmentation of dataset items with probability p, with support for detection and segmentation item types and datasets (@DerrickUnleashed, #381).
• Added item_transform_resize() for resizing dataset items, with support for detection and segmentation item types and datasets (@srishtiii28, #362).
• item_transform_rotate() now supports segmentation items and datasets, rotating the masks alongside the image (@srishtiii28, #379).
• Added item_transform_crop() for cropping dataset items at a specified location and size, with support for detection and segmentation item types and datasets (@DerrickUnleashed, #371).
• Added item_transform_pad() for padding dataset items on all sides, with support for detection, segmentation and rotated-box item types and datasets (@DerrickUnleashed, #373).
• Added item_transform_perspective() for applying a perspective transform to dataset items, with support for detection and segmentation item types and datasets, and implemented transform_perspective() for tensor images (@DerrickUnleashed, #377).
• Added item_transform_random_resize_crop() for cropping dataset items to a random area and aspect ratio before resizing them to a given size, with support for detection, segmentation and rotated-box item types and datasets (@srishtiii28, #386).
• Added item_transform_random_crop() for cropping dataset items at a random location with optional padding, with support for detection and segmentation item types and datasets (@DerrickUnleashed, #387).
• Added item_transform_random_erasing() for randomly erasing a rectangular region of dataset items with probability p, with support for detection, segmentation and rotated-box item types and datasets (#389).
• Added item_transform_random_affine() for applying an affine transformation drawn from the given ranges to dataset items, with support for detection, segmentation and rotated-box item types and datasets (@srishtiii28, #390).
• Added item_transform_random_perspective() for applying a random perspective transformation to dataset items with probability p, with support for detection, segmentation and rotated-box item types and datasets (@DerrickUnleashed, #393).
• Detection datasets (coco_detection_dataset(), pascal_detection_dataset() and the rf100_*_collection()s) now inherit the object_detection_dataset class and their item target y the object_detection_target class. Segmentation datasets (coco_segmentation_dataset(), pascal_segmentation_dataset(), cityscapes_dataset(), oxfordiiitpet_segmentation_dataset() and rf100_peixos_segmentation_dataset()) now inherit the segmentation_dataset class and their item target y the segmentation_target class. Target transforms now dispatch on those classes: target_transform_resize(), target_transform_rotate(), target_transform_affine() and target_transform_sahi_crop() take an object_detection_target, and target_transform_coco_masks() and target_transform_trimap_masks() a segmentation_target. Bare-list targets remain supported and retain their list class (@srishtiii28, #391).
• Added a “Visualization utilities” article covering vision_make_grid(), draw_bounding_boxes(), draw_segmentation_masks() and draw_keypoints() on the output of model_rfdetr_base() and model_fcn_resnet50() (@srishtiii28, #400).
• Added a “torchvision classes” vignette sketching what each class holds: the image_with_bounding_box, image_with_rotated_box and image_with_segmentation_mask items, their object_detection_target and segmentation_target, the sahi_split returned by prepare_sahi_split(), and which transform and drawing functions dispatch on each (@srishtiii28, #396).

Bug fixes and improvements

• draw_bounding_boxes() now keeps labels and colors aligned with their boxes when lazy = TRUE filters out degenerate boxes (#415).
• Fixed box alignment in padded center crops and random crops, and corrected five- and ten-crop geometry (#414).
• model_vit_*() predictions now depend on the input image: attention runs over the patches of each image, the positional embedding is registered so it is loaded from the pretrained weights, and layer norm uses eps = 1e-6 (@sebffischer, #405).
• model_maxvit() is re-ported from PyTorch and now reproduces its predictions with the pretrained weights (@sebffischer, #405).
• draw_bounding_boxes() now accepts degenerate bounding boxes by default with the lazy = TRUE parameter (#400).
• vision_make_grid() now accepts multiple 3D tensors with mixed uint8 and float dtype, while preserving positional scale, num_rows, padding and pad_value arguments (#398). Use per_row for the number of images per row; named num_rows remains supported as a deprecated alias.
• transform_random_affine() now accepts a bare number for shear. It used to widen degrees instead of shear, which left the shear range incomplete and made the sampling fail (#390).
• item_transform_rotate() no longer truncates the rotation angle to a whole number of degrees when the target stores its boxes as an integer tensor, as pascal_detection_dataset() does (#360).
• transform_random_resized_crop() no longer loses the last row and column of the image, and no longer pads with a black edge when it falls back to a central crop (#386).
• transform_crop() now pads the result with zeros when the crop leaves the image, so that the output always has the requested size. It previously returned only the part of the image the crop covered, which could even be empty.
• transform_center_crop() now pads correctly when the requested size is not square. Height and width were used in the wrong order in the padding branch, so e.g. cropping a (16, 20) image to c(24, 30) returned a (22, 0) image.
• transform_rgb_to_grayscale() now returns an image with one channel instead of dropping the channel dimension, i.e. (3, H, W) becomes (1, H, W) and not (H, W). This also fixes transform_grayscale() for batched input.
• hsv2rgb() computed the fractional part of the hue sector and the t component incorrectly, so converting an image to HSV and back did not return the original image. This affected transform_adjust_hue() and transform_color_jitter(hue = ...).
• transform_adjust_hue() now handles negative hue_factors, which produced a black image before, and rejects images that do not have 1 or 3 channels instead of silently dropping the additional ones.
• nms() now uses torchvisionlib::ops_nms() when torchvisionlib is installed, speeding up inference for model_fasterrcnn_*() and model_maskrcnn_*() (#321, #322).
• Lists and vectors are now preallocated to their target size instead of being grown one element at a time (@srishtiii28, #335).
• model_convnext_small_22k1k() now defaults to num_classes = 1000. It is pretrained on Imagenet 22k and fine-tuned on Imagenet 1k, so its published weights have a 1000-class head, and the previous default of 21841 made pretrained = TRUE fail to load the state dict. (@sebffischer, #376)
• model_inception_v3(pretrained = TRUE, aux_logits = FALSE) now actually removes the auxiliary classifier. Previously aux_logits = FALSE was silently ignored for pretrained models, so the network returned a list of both heads instead of a single tensor in training mode. (@sebffischer, #375)
• model_rfdetr_*() now match the reference rf-detr implementation numerically, accept padding masks, and use the official configurations of the small, medium and base variants (@sebffischer, #413).
• model_fasterrcnn_*() and model_maskrcnn_*() inference now reproduces torchvision numerically, supports batches, and loads pretrained weights strictly (@sebffischer, #412).
• model_inception_v3() and model_resnet*() without pretrained weights now initialise all layers as in PyTorch, and zero_init_residual = TRUE takes effect (@sebffischer, #411).
• The progress argument of the model constructors is now honoured (@sebffischer, #411).
• model_convnext_*_upernet() now apply the per-stage layer norms of the pretrained backbone and default to num_classes = 150 when pretrained = TRUE (@sebffischer, #410).
• model_convnext_*_detection() default to num_classes = 90 and pass score_thresh, nms_thresh and detections_per_img to the detection head (@sebffischer, #410).
• model_convnext_*() blocks now apply stochastic depth in training mode (@sebffischer, #410).
• model_efficientnet_*() and model_efficientnet_v2_*() now match torchvision (SiLU squeeze-excitation, batch norm settings), apply stochastic depth in training, and accept dropout without shifting other arguments (@sebffischer, #409).
• model_mobilenet_v3_*() use torchvision’s batch norm settings, gain a norm_layer argument and stay in training mode when pretrained = FALSE; model_mobilenet_v2() handles stages with n = 0 (@sebffischer, #409).
• model_fcn_resnet*() now use a dilated backbone as in PyTorch, fixing the pretrained segmentations, and no longer download the backbone weights when pretrained = TRUE (@sebffischer, #408).
• model_deeplabv3_*() keep the auxiliary classifier when pretrained = TRUE and aux_loss = NULL, as documented (@sebffischer, #408).
• model_lw_detr_*() now match the reference LW-DETR implementation numerically, which strongly improves the pretrained detections (@sebffischer, #407).
• model_facenet_pnet(), model_facenet_rnet(), model_facenet_onet() and model_mtcnn() take the face softmax over the class dimension instead of the batch dimension (@sebffischer, #406).
• model_facenet_inception_resnet_v1() now matches facenet-pytorch (missing repeat_1 stage, Mixed_6a paddings); pretrained = TRUE loads "vggface2" and num_classes defaults to NULL (@sebffischer, #406).

torchvision 0.9.0

Breaking changes

• Split coco_detection_dataset() into coco_detection_dataset() (detection only) and new coco_segmentation_dataset() (instance segmentation) reducing memory usage by ~50% (@Chandraveersingh1717, #280).
• coco_classes() is now aligned with the 90 sparse pytorch COCO classses, in order to match pretrained model predictions. (@318).
• Renamed $categories to $classes for consistency (character vector of class names; old attribute is deprecated with a warning) (#300).

New features

• Added resolution function for COCO Imagenet_21k and Pascal_VOC classes and labels (#284).
• Added article showcasing model_fcn_resnet50() with visualization utilities draw_segmentation_masks() and vision_make_grid() (@DerrickUnleashed, #281).
• Added collection dataset catalog with search_collection(), get_collection_catalog(), and list_collection_datasets() functions for discovering and exploring collections (#271, @ANAMASGARD).
• Added target_transform_coco_masks() and target_transform_trimap_masks() transformation functions for explicit segmentation mask generation (@ANAMASGARD).
• Added support for connectivity argument for drawing lines between keypoints in draw_keypoints() (@DerrickUnleashed, #303).

New models

• Added model_maskrcnn_resnet50_fpn() and model_maskrcnn_resnet50_fpn_v2() for instance segmentation (#278, @ANAMASGARD).
• Added model_convnext_*_detection() for object detection, with * within tiny/small/base (#262, @ANAMASGARD).
• Added model_convnext_*_fcn() and model_convnext_*_upernet() for semantic segmentation, with * within tiny/small/base (#265, @ANAMASGARD).

New datasets

• Added vggface2_dataset() for loading the VGGFace2 dataset (@DerrickUnleashed, #238).
• Added moth dataset to rf100_biology_collection() and currency and wine_label to rf100_document_collection() (#274).

Bug fixes and improvements

• model_maskrcnn_*() and model_fasterrcnn_*() now match the pytorch implementation (#318).
• transform_ now correctly manage batched 4D torch tensors and 4D arrays (#313).
• mnist_datataset() and derivatives now correctly return item x() values with a 1-channel dimension (@Chandraveersingh1717, #307).
• transform_affine(), transform_rotate() and random derivatives now use interpolation and fill parameter in favor of resampleand fillcolor (@Chandraveersingh1717, #299).
• Fixed draw_keypoints() documentation and error message (@srishtiii28, #296).
• Standardized dataset messages: download messages now include split information, success messages show image count and class count for consistency.
• fix model_fasterrcnn_* did not provide boxes output normalized to image size, did not manage batches, fix performance of the roi_align() function (#284).
• fix rf100 collection bounding-box now consider the correct native COCO format being ‘xywh’ (#272).
• Remove .getbatch method from MNIST as it is providing inconsistent tensor dimensions with .getitem due to non-vectorized transform_ operations (#264).
• Added article for draw_keypoints() (@DerrickUnleashed, #303).
• Fix typos and align model documentation for model_deeplabv3_* and model_convnext_*_detection() to ensure consistency (@DerrickUnleashed #302).
• Fixed draw_keypoints() documentation: corrected connectivity parameter type, obsolete note, colors description default (@srishtiii28 #296).

torchvision 0.8.0

New datasets

• Added lfw_people_dataset() and lfw_pairs_dataset() for loading Labelled Faces in the Wild (LFW) datasets (@DerrickUnleashed, #203).
• Added places365_dataset() for loading the Places365 dataset (@koshtiakanksha, #196).
• Added pascal_segmentation_dataset(), and pascal_detection_dataset() for loading the Pascal Visual Object Classes datasets (@DerrickUnleashed, #209).
• Added whoi_plankton_dataset(), whoi_small_plankton_dataset(), and whoi_small_coral_dataset() (@cregouby, #236).
• Added rf100_document_collection(), rf100_medical_collection(), rf100_biology_collection(), rf100_damage_collection(), rf100_infrared_collection(), and rf100_underwater_collection(). These are collections of datasets from Roboflow 100 under the same theme, for a total of 35 datasets (@koshtiakanksha, @cregouby, #239).
• Added rf100_peixos_segmentation_dataset(). (@koshtiakanksha, @cregouby, #250).

New models

• Added model_maxvit() for MaxViT: Multi-Axis Vision Transformer (#229, @koshtiakanksha).
• Added model_facenet_pnet(), model_facenet_rnet(), and model_facenet_onet() for Facenet MTCNN face detection models. (@DerrickUnleashed, #227)
• Added model_mtcnn() and model_inception_resnet_v1() models for face detection and recognition. (@DerrickUnleashed, #217)
• Added model_mobilenet_v3_large() and model_mobilenet_v3_small() models for efficient image classification. (@DerrickUnleashed, #237)
• Added 8 of the model_convnext_() family models for image classification, thanks to @horlar1 contribution. (@cregouby, #251)
• Added 2 model_fasterrcnn_resnet50_() models and 2 model_fasterrcnn_mobilenet_v3_large_() for object detection. (@koshtiakanksha, #251)

New features

• Added imagenet_label() and imagenet_classes() for ImageNet classes resolution (#229, @koshtiakanksha).
• base_loader() now accept URLs (@cregouby, #246).
• draw_segmentation_masks() now accepts semantic segmentation models torch_float() output. (@cregouby #247)
• MNIST datasets and Roboflow 100 collections now have a .getbatch attached method (@cregouby #255)

Bug fixes and improvements

• Switch pre 0.5.0 models to their /v2/ URL in torch-cdn.mlverse.org. (#215)
• Models are now separated in the documentation by tasks between classification, object detection, and semantic segmentation models (@cregouby, #247)
• Breaking Change : Refactoring of coco_* dataset family now provides each item$x being an image array (for consistency with other datasets). You can use transform = transform_to_tensor to restore the previous x output to be a torch_tensor().
• transform_ are now documented into 3 different categories: unitary transformations, random transformations and combining transformations. (@cregouby, #250)
• Deprecation : emnist_dataset is deprecated in favor of emnist_collection() (@cregouby, #260).

torchvision 0.7.0

New datasets

• Added fashion_mnist_dataset() for loading the Fashion-MNIST dataset (@koshtiakanksha, #148).
• Added eurosat_dataset(), eurosat_all_bands_dataset(), and eurosat100_dataset() for loading RGB, all-band, and small-subset variants of the EuroSAT dataset (@cregouby, #126).
• Added qmnist_dataset() for loading the QMNIST dataset (@DerrickUnleashed, #153).
• Added emnist_dataset() for loading the EMNIST dataset (@DerrickUnleashed, #152).
• Added fgvc_aircraft_dataset() for loading the FGVC-Aircraft dataset (@DerrickUnleashed, #156).
• Added coco_detection_dataset() and coco_caption_dataset() for loading the MS COCO detection and captions datasets (@koshtiakanksha, #161, #172).
• Added caltech101_dataset() and caltech256_dataset() for loading the Caltech 101 and 256 datasets (@DerrickUnleashed, #158).
• Added fer_dataset() for loading the FER-2013 dataset (@DerrickUnleashed, #154).
• Added flowers102_dataset() for loading the Flowers102 dataset (@DerrickUnleashed, #157).
• Added flickr8k_dataset() and flickr30k_dataset() for loading the Flickr8k and Flickr30k datasets (@DerrickUnleashed, #159).
• Added oxfordiiitpet_dataset(), oxfordiiitpet_binary_dataset(), and oxfordiiitpet_segmentation_dataset() for loading the Oxford-IIIT Pet datasets (@DerrickUnleashed, #162).
• Added rf100_document_collection(), rf100_underwater_collection(), rf100_medical_collection(), rf100_biology_collection(), and rf100_peixos_segmentation_dataset() for loading Roboflow 100 datasets (@koshtiakanksha, #239).

New models

• Added EfficientNet model family (B0–B7) – scalable CNNs for image classification. (#166, @koshtiakanksha)
• Added EfficientNetV2 model family (V2-S/M/L) – improved EfficientNet models for faster training. (#166, @koshtiakanksha)
• Added model_vit_b_16(), model_vit_b_32(), model_vit_l_16(), model_vit_l_32(), and model_vit_h_14() for loading Vision Transformer models (@DerrickUnleashed, #202).

New features

• tensor_image_display() and tensor_image_browse() now accept all tensor_image dtypes (@cregouby, #115).
• draw_bounding_boxes() and draw_segmentation_masks() now accept image_with_bounding_box and image_with_segmentation_mask inputs which are the default items class for respectively detection datasets and segmentation datasets (@koshtiakanksha, #175).
• fgvc_aircraft_dataset() gains support for annotation_level = "all" (@DerrickUnleashed, #168).
• folder_dataset() now supports TIFF image formats (@cregouby, #169).
• New nms() and batched_nms() functions provide Non-Maximum Suppression utilities. Added box_convert() to convert between bounding box formats (@Athospd, #40).

Minor bug fixes and improvements

• transform_rotation() now correctly uses width × height for image size instead of width × width (@cregouby, #114).
• Clarified documentation for transform_affine() to reduce confusion with transform_random_affine() (@cregouby, #116).
• Added French translations for message outputs (@cregouby, #112).

torchvision 0.6.0

• Remove again dependency on zip::unzip added in version 0.4.0. (#89)
• Improve performance on tinyimagenet-alexnet example (#90, @statist-bhfz)
• Updated URL of downloaded resources to use the new torch CDN. (#109)

torchvision 0.5.1

• Remove usage of torch_lstsq that was removed in torch v0.10.0

torchvision 0.5.0

• Bugs fixed in transform_adjust_hue() and transform_linear_transformation() (#72, #73, @sebffischer)
• add draw_bounding_boxes() , draw_segmentation_masks() and draw_keypoints() on top of image tensors, and add a convenience tensor_image_browse() and tensor_image_display() functions to visualize image tensors respectively in browser or in X11 device (#80, @cregouby)
• Added the InceptionV3 model. (#82)

torchvision 0.4.1

• Implemented MobileNetV2 (#60)
• Improved vignettes so they use nnf_cross_entropy for numerical stability. (#61)
• Implement the full list of ResNet model family (#66, @cregouby)
• Improved how datasets and models are downloaded by using a large timeout by default and downloading to temporary file to avoid hard to debug errors when the files are corrupt. (#67)

torchvision 0.4.0

• Added a dependency on zip to zip::unzip the tinyimagenet dataset.
• Removed all usages of torch::enumerate() from docs and tests in favor of coro::loop() (#57)
• Fixed non-namespaced calls to torch. (#58)

torchvision 0.3.0

• Use a self hosted version of the MNIST dataset to avoid frequent download failures. (#48)
• Fix torch_arange calls after breaking change in torch. (#47)
• Fix bug in transform_resize when passing size with length 1. (#49)

torchvision 0.2.0

• Fixed bugs in transform_rotate. (#31)
• Fixed bugs in transform_random_affine and transform_affine (#32)
• Added VGG model (#35)

torchvision 0.1.0

• Added a NEWS.md file to track changes to the package.