The Unconstrained Visible Spectrum Contact Lens Iris (UVCLI) database consists of 3,802 iris images captured in visible spectrum at two locations: indoors (with controlled illumination) and outdoors (with varying environmental scenarios) in two sessions. These images have been captured one at a time using EOS 60D DSLR Canon camera with EFS 60mm f/2.8 Macro USM Fixed lens. In each session of data acquisition, a minimum of six images of each eye when wearing textured contact lens and a minimum of six images of each eye without any lens (real) are acquired. These images are captured indoors and this process is repeated in the outdoor environment. Outdoor images have been collected at varying times of the day (such as afternoon and night) and varying weather conditions (such as sunny and cloudy). Majority of the first session outdoor data collection is conducted during the daytime.
The database can be downloaded from the following link:
Unconstrained Visible Spectrum Contact Lens Iris (UVCLI) (4.40GB)
- To obtain the password for the compressed file, email the duly filled license agreement to databases@iab-rubric.org with the subject line "License agreement for Unconstrained Visible Spectrum Contact Lens Iris (UVCLI)"
NOTE: The license agreement has to be signed by someone having the legal authority to sign on behalf of the institute, such as the head of the institution or registrar. If a license agreement is signed by someone else, it will not be processed further.
This database is available only for research and educational purpose and not for any commercial use. If you use the database in any publications or reports, you must refer to the following paper: - D. Yadav, N. Kohli, M. Vatsa, R. Singh, & A. Noore, Unconstrained visible spectrum iris with textured contact lens variations: Database and benchmarking, In IEEE International Joint Conference on Biometrics (IJCB), pages 574-580, 2017.
The WVU Unconstrained Multi-sensor Iris Presentation Attack (UnMIPA) database consists of 18,706 iris images from 81 subjects. It contains textured contact lenses of blue, green, gray, violet, and brown colors from the following brands: Bausch and Lomb, Freshlook Dailies, Freshlook Colorblends, and Celebration. For each subject, iris images are acquired using 3 iris scanners: CMITECH EMX-30, IriShield BK 2121U, and IriShield MK 2120U.
For each subject in the database, iris images have been captured using different cameras with and without textured contact lens in both indoors (controlled) and outdoors (unconstrained) environment. The iris images acquired outdoor have been acquired at different times of the day as well as in different weather conditions.
The database can be downloaded from the following link:
- To obtain the password for the compressed file, email the duly filled license agreement to databases@iab-rubric.org with the subject line "License agreement for WVU UnMIPA Database"
NOTE: The license agreement has to be signed by someone having the legal authority to sign on behalf of the institute, such as the head of the institution or registrar. If a license agreement is signed by someone else, it will not be processed further.
This database is available only for research and educational purpose and not for any commercial use. If you use the database in any publications or reports, you must refer to the following paper: - D. Yadav, N. Kohli, M. Vatsa, R. Singh, and A. Noore. Detecting Textured Iris Lens in Uncontrolled Environment using DensePAD, In IEEE Conference on Computer Vision and Pattern Recognition Workshop on Biometrics, 2019.
IIITD-WVU dataset of LivDet Iris 2017 competition has been designed as a cross-database evaluation where the sensors as well as the acquisition environments for the training and testing sets are different. It has the following two partitions:
1) Train: The training set of IIITD-WVU dataset is developed using 2,250 real and 1,000 textured contact lens iris images from IIIT-Delhi Contact Lens Iris (CLI) database which is captured in controlled settings. Also, 3,000 print attack images are selected from IIITD Iris Spoofing (IIS) database. It contains 4 folders: Contacts (containing textured contact lens iris images), Printouts (containing print attack images of live/real iris images), Contacts+Printouts (containing print attack images of contact lens iris images), and Live (containing live/real iris images).
2) Test: It is a novel multi-session iris presentation attack detection dataset comprising 4,209 iris images. The images in this dataset are captured using IriShield MK2120U mobile iris sensor at two different locations: indoors (controlled illumination) and outdoors (varying environmental situations). In the two sessions of data collection, iris images are acquired one at a time while wearing textured/patterned contact lens and without any lens (real). It contains 4 folders: Contacts (containing textured contact lens iris images), Printouts (containing print attack images of live/real iris images), Contacts+Printouts (containing print attack images of contact lens iris images), and Live (containing live/real iris images).
The database can be downloaded from the following link: IIIT-WVU subset of LivDet 2017 (1.81GB)
(CRC32: CRC-32: 55c27972, MD5: f1bad2ed4500099eb4fbc3db1c0100a2, SHA-1: 2005947639eb29a4b5ebcab1af067412a1e525f0)
To obtain the password for the compressed file, email the duly filled license agreement to databases@iab-rubric.org and afzel.noore@mail.wvu.edu with the subject line "License agreement for WVU Mobile Iris Spoofing Dataset". This database is available only for research and educational purpose and not for any commercial use.
NOTE: The license agreement has to be signed by someone having the legal authority to sign on behalf of the institute, such as the head of the institution or registrar and send to both Email IDs. If a license agreement is signed by someone else, it will not be processed further.
Children Multimodal Biometric Database (CMBD) consists of iris, fingerprint, and face images of over 100 children (age range of 18 months to 4 years), acquired over two sessions. The iris images are captured using Cross-Match iris scanner, fingerprints are captured using Cross-Match L-scan slap fingerprint scanner and face images are captured using Nikon D90 DSLR. It has data for 106 across both the session and for all 3 modalities. However, to protect the identities of the participants, we have provided iris and fingerprint images only.
- The database can be downloaded from the following link.
Children Multimodal Biometric Database (CMBD) (760MB)
- To obtain the password for the compressed file, email the duly filled license agreement to databases@iab-rubric.org with the subject line "License agreement for Children Multimodal Biometric Database (CMBD) Database"
NOTE: The license agreement has to be signed by someone having the legal authority to sign on behalf of the institute, such as the head of the institution or registrar. If a license agreement is signed by someone else, it will not be processed further.
This database is available only for research and educational purpose and not for any commercial use. If you use the database in any publications or reports, you must refer to the following paper: - P. Basak, S. De, M. Agarwal, A. Malhotra, R. Singh, and M. Vatsa. Multimodal Biometric Recognition for Toddlers and Pre-School Children, In IEEE International Joint Conference On Biometrics, 2017..
Cataract Mobile Periocular Database (CMPD) contains periocular images of pre-surgery and post-surgery cataract patients. In pre-post surgery, 145 and 99 subjects are present. However, 56 subjects are common in pre, and post set. The database contains 2380 number of samples of 4608 x 3456 resolution. These images are captured via a mobile device which has 16-megapixel camera resolution.
The database can be downloaded from the following link:
Cataract Mobile Periocular Database (CMPD) (5.23 GB) (CRC-32: b7e80785, MD5: 68c274ba86d65472459a48ea85eb8a0b)
- To obtain the password for the compressed file, email the duly filled license agreement to databases@iab-rubric.org with the subject line "License agreement for Cataract Mobile Periocular Database (CMPD)"
NOTE: The license agreement has to be signed by someone having the legal authority to sign on behalf of the institute, such as the head of the institution or registrar. If a license agreement is signed by someone else, it will not be processed further.
This database is available only for research and educational purpose and not for any commercial use. If you use the database in any publications or reports, you must refer to the following paper: - R. Keshari, S. Ghosh, A. Agarwal, R. Singh, M. Vatsa, Mobile Periocular Matching with Pre-Post Cataract Surgery, IEEE International Conference on Image Processing, 2016.
Healthy Samples & Segmentation Masks
- To obtain the password for the compressed file, email the duly filled license agreement to databases@iab-rubric.org with the subject line "License agreement for Healthy Samples and Annotations (HSA)"
NOTE: The license agreement has to be signed by someone having the legal authority to sign on behalf of the institute, such as the head of the institution or registrar. If a license agreement is signed by someone else, it will not be processed further.
This database is available only for research and educational purpose and not for any commercial use. If you use the database in any publications or reports, you must refer to the following paper: - R. Keshari, S. Ghosh, A. Agarwal, R. Singh, M. Vatsa, Mobile Periocular Matching with Pre-Post Cataract Surgery, IEEE International Conference on Image Processing, 2016
- Khurshid, M., Akhter, Y., Singh, R. and Vatsa, M., “AssistDistil for Medical Image Segmentation” (venue to be updated)
IIIT-D CLI database comprises of 6570 iris images pertaining to 101 subjects. Both left and right iris images of each subject are captured and therefore, there are 202 iris classes. The lenses used in the database are soft lenses manufactured by either CIBA Vision or Bausch and Lomb. For textured lenses, four colors are used. To study the effect of acquisition device on contact lenses, iris images are captured using two iris sensors: (1) Cogent dual iris sensor (CIS 202) and (2) VistaFA2E single iris sensor. The database contains a minimum of three images for each iris class in each of the above mentioned lens categories for both the iris. To perform the experiments for lens detection, images pertaining to the first 50 subjects are used for training and the remaining 51 subjects are used for testing.
The database can be downloaded from the following link: IIITD CLI Database (1.1GB)
(CRC32: 19770433, MD5: BA3711CE6E997FF37E6788E2DE774A12, SHA-1: BDDD6E058520499B60D8A8925E23C9931F338289)
To obtain the password for the compressed file, email the duly filled license agreement to databases@iab-rubric.org with the subject line "License agreement for IIIT-D CLI Database". This database is available only for research and educational purpose and not for any commercial use.
NOTE: The license agreement has to be signed by someone having the legal authority to sign on behalf of the institute, such as the head of the institution or registrar. If a license agreement is signed by someone else, it will not be processed further.
Publications
- D. Yadav, N. Kohli, J. S. Doyle, R. Singh, M. Vatsa, and K. W. Bowyer, Unraveling the Effect of Textured Contact Lenses on Iris Recognition, IEEE Transactions on Information Forensics and Security, 2014 (Accepted).
- D. Yadav, N. Kohli, R. Singh, and M. Vatsa, Revisiting Iris Recognition with Color Cosmetic Contact Lenses, 6th IAPR International Conference on Biometrics, June, 2013.
The multi-spectral periocular database used in this research comprises images from three spectrum:
- Visible Spectrum: It comprises a total of 310 images correspond to 62 subjects in visible spectrum. Each subject has 5 images corresponding to it.
- Night Vision: It comprises a total of 310 images correspond to 62 subjects in Night Vision mode. Each subject has 5 images corresponding to it.
- Near Infrared: It comprises a total of 620 images correspond to 62 subjects in Near Infrared spectrum. Each subject has 10 images corresponding to it. 5 images are of Right eye and 5 images are of Left eye.
The database can be downloaded from the following link:
IMP Database (41 MB) (CRC32: F3ED9F69, MD5: 0437F98FA8E425E42DC435E420EEB2FC, SHA-1: 1B9C93966B56959C81A6242B2BAAE4125421E8EC)
- To obtain the password for the compressed file, email the duly filled license agreement to databases@iab-rubric.org with the subject line "License agreement for IIIT-D Multispectral Periocular Database"
NOTE: The license agreement has to be signed by someone having the legal authority to sign on behalf of the institute, such as the head of the institution or registrar. If a license agreement is signed by someone else, it will not be processed further.
This database is available only for research and educational purpose and not for any commercial use. If you use the database in any publications or reports, you must refer to the following paper: - A. Sharma, S. Verma, M. Vatsa, and R. Singh On Cross Spectral Periocular Recognition, In Proceedings of International Conference on Image Processing, 2014.
Combined spoofing database (CSD) consists of images from multiple publicly available spoofing databases. It consists of IIIT-Delhi Contact Lens Iris (CLI) Database, IIITD Iris Spoofing (IIS) Database, and Multi-sensor Iris Database. For experimental purposes, independently download Synthetic Database and IIT Delhi Iris Database to include in CSD.
The database can be downloaded from the following link: Iris Combined Spoofing Database (3.25GB)
(MD5: 4880E12A77CD494650688EEF592747DB, SHA-1: 982A10E3BA12DAE45E2CA3390D56354E1738E674)
To obtain the password for the compressed file, email the duly filled license agreement to databases@iab-rubric.org with the subject line "License agreement for Iris Combined Spoofing Database". This database is available only for research and educational purpose and not for any commercial use.
NOTE: The license agreement has to be signed by someone having the legal authority to sign on behalf of the institute, such as the head of the institution or registrar. If a license agreement is signed by someone else, it will not be processed further.
Publications
- The publications details are: N. Kohli, D. Yadav, M. Vatsa, R. Singh, and A. Noore, Detecting Medley of Iris Spoofing Attacks using DESIST , In Proceedings of International Conference on Biometrics: Theory, Applications, and Systems, 2016.
