Container mainly used for lists of torch tensors. Extends lists with pytorch functionality.

List of sequences. Supports the addition operator to concatenate sequence lists.

Class for the sequence in an evaluation.

Container mainly used for dicts of torch tensors. Extends OrderedDict with pytorch functionality.

GOT-10k dataset. Publication: GOT-10k: A Large High-Diversity Benchmark for Generic Object Tracking in the Wild Lianghua Huang,

Wraps the tracker for evaluation and running purposes. args: name: Name of tracking method. parameter_name: Name of parameter file

Computes and stores the average and current value

Data loader. Combines a dataset and a sampler, and provides single- or multi-process iterators over the dataset. Note: The only differen

Track a series of values and provide access to smoothed values over a window or the global series average.

Corner Predictor module

BatchNorm2d where the batch statistics and the affine parameters are fixed. Copy-paste from torchvision.misc.ops with added eps before rq

Imagenet VID dataset. Publication: ImageNet Large Scale Visual Recognition Challenge Olga Russakovsky, Jia Deng, Hao Su, Jonatha

Imagenet VID dataset. Publication: ImageNet Large Scale Visual Recognition Challenge Olga Russakovsky, Jia Deng, Hao Su, Jonatha

LaSOT dataset. Publication: LaSOT: A High-quality Benchmark for Large-scale Single Object Tracking Heng Fan, Liting Lin, Fan Yan

Very simple multi-layer perceptron (also called FFN)

The COCO dataset. COCO is an image dataset. Thus, we treat each image as a sequence of length 1. Publication: Microsoft COCO: Common Obj

The COCO dataset. COCO is an image dataset. Thus, we treat each image as a sequence of length 1. Publication: Microsoft COCO: Common Obj

This is the base class for OSTrack

Actor for training OSTrack models

2D Image to Patch Embedding

Class for tracker parameters.

TrackingNet dataset. Publication: TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild. Matthias Mue

TrackingNet dataset. Publication: TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild. Matthias Mue

Vision Transformer A PyTorch impl of : `An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale` - https://arxiv.org

Vision Transformer with candidate elimination (CE) module A PyTorch impl of : `An Image is Worth 16x16 Words: Transformers for Image Recogniti

Base class for actor. The actor class handles the passing of the data through the network and calculation the loss

Base class for all datasets.

Base class for image datasets

Base class for Processing. Processing class is used to process the data returned by a dataset, before passing it through the network. For exampl

Base class for all trackers.

Base trainer class. Contains functions for training and saving/loading checkpoints. Trainer classes should inherit from this one and overload the

Base class for video datasets

Class for feature specific parameters

GOT-10k dataset. Publication: GOT-10k: A Large High-Diversity Benchmark for Generic Object Tracking in the Wild Lianghua Huang,

NUS-PRO dataset

Loss that uses a 'hinge' on the lower bound. This means that for samples with a label value smaller than the threshold, the loss is zero if the p

LaSOT test set consisting of 280 videos (see Protocol-II in the LaSOT paper) Publication: LaSOT: A High-quality Benchmark for Large-

LaSOT test set consisting of 280 videos (see Protocol-II in the LaSOT paper) Publication: LaSOT: A High-quality Large-scale Single

LaSOT test set consisting of 280 videos (see Protocol-II in the LaSOT paper) Publication: LaSOT: A High-quality Benchmark for Large-

The COCO object detection dataset. Publication: Microsoft COCO: Common Objects in Context. Tsung-Yi Lin, Michael Maire, Serge J.

NFS dataset. Publication: Need for Speed: A Benchmark for Higher Frame Rate Object Tracking H. Kiani Galoogahi, A. Fagg, C. Huang

Normalize image

OTB-2015 dataset Publication: Object Tracking Benchmark Wu, Yi, Jongwoo Lim, and Ming-hsuan Yan TPAMI, 2015 http:

Horizontally flip image randomly with a probability p.

Horizontally flip image randomly with a probability p. The difference is that the coord is normalized to [0,1]

The processing class used for training LittleBoy. The images are processed in the following way. First, the target bounding box is jittered by ad

Training settings, e.g. the paths to datasets and networks.

TC-128 Dataset (78 newly added sequences) modified from the implementation in got10k-toolkit (https://github.com/got-10k/toolkit)

TC-128 Dataset modified from the implementation in got10k-toolkit (https://github.com/got-10k/toolkit)

TNL2k test set

Converts image to BGR

Converts image to grayscale with probability

Convert to a Tensor

Convert to a Tensor and jitter brightness

TrackingNet test set. Publication: TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild. Matthias Mu

Class responsible for sampling frames from training sequences to form batches. The sampling is done in the following ways. First a dataset is s

A set of transformations, used for e.g. data augmentation. Args of constructor: transforms: An arbitrary number of transformations, derive

Base class for transformation objects. See the Transform class for details.

UAV123 dataset. Publication: A Benchmark and Simulator for UAV Tracking. Matthias Mueller, Neil Smith and Bernard Ghanem

VOT2018 dataset Publication: The sixth Visual Object Tracking VOT2018 challenge results. Matej Kristan, Ales Leonardis, Jiri