The Eighth Visual Object Tracking VOT2020 Challenge Results

By Matej Kristan, Ales Leonardis, Jiri Matas, Michael Felsberg, Roman P. Pflugfelder, Joni-Kristian Kämäräinen, Martin Danelljan, L. Č. Zajc, Alan Lukezic, O. Drbohlav and 97 more. 2020

345 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

In Evaluation & Benchmarks

Citations in the corpus

Visual Prompt Multi-Modal Tracking (2023, methodology): “CA3DMS [31] SiamM Ds [18] Siam LTD [17] LTDSEd [18] DAL [37] CLGS D [17] LTMU B [17] ATCAIS [17] DDiMP [17] DeT [47] OSTrack [49] SPT [60] ProTrack [48] ViPT [47].”
ODTrack: Online Dense Temporal Token Learning for Visual Tracking (2024, methodology): “VOT2020.”
MixFormer: End-to-End Tracking with Iterative Mixed Attention (2022, methodology): “• The proposed MixFormer sets a new state-of-the-art performance on five challenging benchmarks, including VOT2020 [26], LaSOT [17], TrackingNet [41], GOT-10k [23], and UAV123 [40].”
CrossFuse: A novel cross attention mechanism based infrared and visible image fusion approach (2024, methodology): “The test images are selected from TNO [62] and VOT-RGBT [63], comprising 21 and 40 pairs of infrared and visible images, respectively.”
Is First Person Vision Challenging for Object Tracking? (2020, positive): “For the implementation of the multi-start evaluation (MSE) protocol, we followed the details given in [47].”
Transforming Model Prediction for Tracking (2022, methodology): “Comparison to the state of the art of bounding box only methods on VOT2020ST [35] in terms of EAO score.”
Learning Target Candidate Association to Keep Track of What Not to Track (2021, methodology): “So do KeepTrack and KeepTrackFast that perform similarly to the top methods, with a 0.8% and 1.1% AUC gain over the SuperDiMP baseline.”
RFN-Nest: An end-to-end residual fusion network for infrared and visible images (2021, methodology): “In VOT2020 [45], the video sequences are the same as VOT2019 [54], but a new performance evaluation protocol is introduced for short-term tracker evaluation (includes RGBT sub-challenge).”
Learning Spatio-Temporal Transformer for Visual Tracking (2021, methodology): “Different from previous reset-based evaluations [23], VOT2020 [22] proposes a new anchor-based evaluation protocol and uses binary segmentation masks as the groundtruth.”
SeqTrack: Sequence to Sequence Learning for Visual Object Tracking (2023, methodology): “VOT2020 [25] benchmark contains 60 challenging videos.”
STMTrack: Template-free Visual Tracking with Space-time Memory Networks (2021, methodology): “[19], that uses a non-local block to augment the response map generated by a pixel-wise correlation since the longerrange dependencies can produce more precise target boundary decision information.”
Track Anything: Segment Anything Meets Videos (2023, background): “On the other hand, tracking or segmenting in videos faces challenges from scale variation, target deformation, motion blur, camera motion, similar objects, and so on [4, 8, 10, 7, 9, 6].”