PETS2010: Dataset and Challenge
By James M. Ferryman, A. Ellis. 2010
104 citations recorded by the source corpus.
Subfields: Artificial Intelligence, 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 Artificial Intelligence
- Bernt Schiele
- L. Leal-Taix'e
- Konrad Schindler
- Anton Milan
- Ian David Reid
- Rodrigo Benenson
- Vishal M. Patel
- Stefan Roth
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Bernt Schiele
- Ming-Hsuan Yang
- Xiaogang Wang
- Jitendra Malik
- Dacheng Tao
- Trevor Darrell
- Andrew Zisserman
In Evaluation & Benchmarks
- L. Leal-Taix'e
- Konrad Schindler
- Ian David Reid
- Bernt Schiele
- Anton Milan
- Daniel Cremers
- Rodrigo Benenson
- Luc Van Gool
Citations in the corpus
- Learning Non-maximum Suppression (2017, background): “We report both single-(PETS pedestrians) and multi-class results (COCO dataset), both showing improvements over Gree-dyNMS.”
- MOT16: A Benchmark for Multi-Object Tracking (2016, background): “One of the few exceptions is the well known PETS dataset [16], targeted primarily at surveillance applications.”
- Multi-source Multi-scale Counting in Extremely Dense Crowd Images (2013, methodology): “, USCD dataset with density of 11 − 46 people per frame [4], Mall dataset with density of 13−53 individuals per frame [5], and PETS dataset containing 3 − 40 people per frame [9].”
- Multiple Hypothesis Tracking Revisited (2015, methodology): “We use 11 MOT Challenge [24] training sequences and 5 PETS 2009 [14] sequences for these experiments.”
- CrowdNet: A Deep Convolutional Network for Dense Crowd Counting (2016, background): “Some works in the crowd counting literature experiment on datasets having sparse crowd scenes [2, 10], such as UCSD dataset [2], Mall dataset [3] and PETS dataset [7].”
- A Convnet for Non-maximum Suppression (2015, methodology): “PETS We use 8 of the PETS sequences (Ferryman & Ellis, 2010), ∼200 frames each, that we split in 5 for training (S1L1-1, S1L1-2, S1L2-1, S1L2-2, S2L1, and S3MF1), 1 for validation (S2L3) and 1 for testing (S2L2).”
- MOTChallenge 2015: Towards a Benchmark for Multi-Target Tracking (2015, methodology): “While some annotations follow pedestrians even under full occlusion [20], others create a new trajectory once the pedestrian reappears [18].”
- MOTChallenge: A Benchmark for Single-Camera Multiple Target Tracking (2020, methodology): “Typically, methods are tuned for each sequence, reaching over 90% accuracy in well-known sequences like PETS (Ferryman and Ellis 2010).”
- Continuous Energy Minimization for Multitarget Tracking (2014, background): “Six of them are part of the PETS 2009/2010 benchmark [17], [18].”