10th Sclera Segmentation and Benchmarking Competition (SSBC 2026)

How to participate?

Registration for the competition can be done via email. If you would like to register and receive the training datasets, please send an email to abhijit.das@hyderabad.bits-pilani.ac.in, dhruv.p10@ahduni.edu.in, and matej.vitek@fri.uni-lj.si with the subject line "SSBC 2026 registration" and with the following information:

Name, Affiliation, Email, Phone number, CV , and Mailing Address.

Additionally, please note which track you would like to participate in (Track 1 on Foundation Models, Track 2 on Weakly Supervised Learning, or both) and attach all the filled out and signed forms for the training data:

• SBVPI 

• MASD

• SynCROI

• MOBIUS  (Track 2 only)

The forms can be downloaded via File -> Download (no extra sharing permissions required).

Execution

 SSBC 2026 will feature two distinct tracks to address divergent trends in the field:

• Track 1: Foundation Models and VLMs. Participants will submit solutions that integrate large multi-modal models into biometric segmentation pipelines. We specifically encourage grounding approaches, in which textual or auxiliary prompts guide the segmentation, or novel architectures built on top of FMs/VLMs. The goal is to leverage semantic understanding while ensuring pixel-level biometric precision.

• Track 2: Label-Efficient Sclera Segmentation. This track focuses on semi-supervised, weakly-supervised, and self-supervised learning. Participants must design models capable of learning from incomplete, sparse, or no labels. This track benchmarks how well algorithms can perform when high-quality annotations are scarce, simulating real-world medical/biometric deployment conditions.

The participants may choose to submit to either track (or both) – submission to both is not required.

To ensure a fair competition and reproducibility of the results, the participants will be expected to provide their code and models at the end of the competition.

Task details

Track 1: Foundation Models & VLMs

• Goal: Submit algorithms that utilize Foundation Models or VLMs as a core component.

• Focus: Architectural novelty (grounding mechanisms, adapter modules, hybrid networks).

• Comparison: Results will be compared against simple lightweight segmentation models provided by organizers.

Track 2: Label-Efficient Learning

• Goal: Submit algorithms trained using label-efficient protocols (semi/weak/self-supervised).

• Constraint: Submissions should focus on learning from limited data.

All participants, whose submission yield competitive performance relative to the simplistic baselines, will be invited to co-author the summary paper of the competition. The summary paper will be submitted to IEEE IJCB 2026 and, if accepted, will appear in IEEE Xplore.

A detailed document with instructions on the competition is available HERE.

Organizing Institutions

Relevant Publications

• M. Vitek, A.Das et al., "Exploring Bias in Sclera Segmentation Models: A Group Evaluation Approach," in IEEE Transactions on Information Forensics and Security, vol. 18, pp. 190-205, 2023. [PDF]

• D. Tomašević; P. Peer, V. Štruc, BiOcularGAN: Bimodal Synthesis and Annotation of Ocular Images, IJCB 2022, pp. 1-10, 2022. [PDF]

• M. Vitek, A. Das et al. , SSBC 2020: Sclera Segmentation Benchmarking Competition in the Mobile Environment, IJCB 2020. [PDF]

• A. Das, U Pal, M. Blumenstein, C. Wang, Y. He, Y. Zhu, Z. Sun, Sclera Segmentation Benchmarking Competition in Cross-resolution Environment, ICB 2019. [PDF]