Ongoing/Completed

I-SIP Lab Datasets

Research Datasets

image hover

Lotus Benchmark

image hover

Blood Pressure Dataset

image hover

EEG Dataset

image hover

IoT-TD Dataset

Lotus Benchmark Dataset

The goal of the Lung-Originated Tumor Segmentation from Computed Tomography Scan (LOTUS) Benchmark is to provide a unique dataset and pre-defined metrics, so that different researchers can develop and evaluate their methods. All the algorithms, proposed during this benchmark, are based on deep learning networks, combined with a false positive reduction technique. Methods developed by the three finalists show promising results in tumor segmentation. However, more effort should be put into reducing the false positive rate. This paper presents an overview of the challenge, along with the proposed algorithms and results.

The goal of the LOTUS benchmark is to provide an opportunity to develop automatic lung tumor segmentation methods based on a unique dataset. The LOTUS benchmark was constructed through the 2018 Video & Image Processing (VIP) Cup competition, which was organized in conjunction with 2018 IEEE International Conference on Image Processing(ICIP) to test and evaluate different algorithms against predefined measures.

VIP-Cup Page at IEEE SPS
VIP-Cup Page at ICIP'18 Conference
2018 VIP-Cup Web Page

Lotus Dataset Link

Lotus Benchmark data can be accessed through the following link:

Lotus Benchmark

On-line Evaluation Link

Online evaluation is on the way, in the meantime to receive your results based on the test-dataset please contact us directly via an email to   arash.mohammadi@concordia.ca

Citation

Please us the following citation to refer to Lotus Dataset

2019
item thumbnail

Lung-Originated Tumor Segmentation from Computed Tomography Scan (LOTUS) Benchmark,

arXiv, 2019 P. Afshar, A. Mohammadi, K.N. Plataniotis, K. Farahani, J.S. Kirby, A. Oikonomou, A. Asif, L. Wee, A. Dekker, X. Wu, et al.
2019
item thumbnail

Lung Cancer Radiomics: Highlights from the IEEE Video and Image Processing Cup 2018 Student Competition [SP Competitions],

IEEE Signal Processing Magazine, vol. 36, no. 1, pp 164-173, 2019 A. Mohammadi, P. Afshar, A. Asif, K. Farahani, J.S. Kirby, A. Oikonomou, K.N. Plataniotis
2015
item thumbnail

Data From NSCLC-Radiomics,

The Cancer Imaging Archive, 2015 H. J.W. L. Aerts, E. R. Velazque, R. T. H. Leijenaar, C. Parmar, P. Grossmann, S. Carvalho, P. Lambin
image hover

Radiomics Pipeline

image hover

2018 VIP-Cup

2018 Video and Image Processing (VIP) Cup, as Part of ICIP'18 Conference.
image hover

Lung Cancer Radiomics

Top

IoT-TD Benchmark Dataset

The importance of availability of benchmark datasets with exact ground truth for BLE-based indoor tracking is an undeniable fact that never can be neglected. Lack of such a BLE-based dataset with ground truth is an obstacle for research reproducibility because only through availability of a benchmark dataset different algorithms can be evaluated and compared against the ground truth (actual labels).

Capitalizing on the advantages of benchmarks, especially given the surge of interest on IoT location-based services where accurate solutions are of significant importance, this dataset is the first step towards achieving this goal to construct a unique dataset, referred to as the IoT-TD benchmark, for advancement of BLE-based indoor tracking/localiztion algorithms. What makes the introduced dataset unique is availability of the “Ground Truth Trajectories” synchronized with the RSSI values collected in a multi-sensor setting together with IMU sensor measurements obtained synchronously from the moving target’s hand-held device. All three components of the dataset are time-stamped and preprocessed being available publicly for future BLE and PDR tracking algorithmic developments.

IoT-TD Dataset Link

IoT-TD Benchmark data can be accessed through the following link, for acquiring password for the dataset please send an email to  arash.mohammadi@concordia.ca

IoT-TD Benchmark
Top