Call for Papers

Scope

ICIAP 2027 is the 24th edition of a series of conferences organised biennially by CVPL, the Italian Member Society of the International Association for Pattern Recognition (IAPR). The conference will take place in Modena, Italy, from 13 to 17 September 2027: workshops and tutorials on 13–14 September, and the main conference on 15–17 September.

ICIAP 2027 invites high-quality original contributions on all aspects of image analysis and processing, computer vision, pattern recognition, machine learning, artificial intelligence, and agentic AI. Both theoretical advances and application-driven work are welcome, as are contributions that bridge the two.

Papers must be original, unpublished and not under review elsewhere at the time of submission. Each accepted paper is expected to be presented at the conference by one of its authors, and at least one author must complete a full registration.

ICIAP 2027 is meant to be a lively forum where researchers can meet, present their work and exchange ideas. The technical program combines oral and poster sessions with keynote lectures by invited, distinguished speakers, and is complemented by satellite workshops and tutorials.

Tracks and topics

ICIAP 2027 is organised in 11 tracks, each handled by a group of area chairs. Topics of interest include, but are not limited to:

Pattern Recognition and Machine Learning + Deep Learning

• Deep learning architectures, representation learning and foundation models for vision
• Self-, semi- and unsupervised learning; transfer, few-shot, continual and long-tail learning
• Generative models for images and video
• Pattern recognition, classification, statistical learning and optimization methods
• Fairness, robustness, explainability and trustworthy vision
• Datasets, benchmarks and evaluation methodology

Image Analysis, Detection, and Recognition

• Image enhancement, restoration and computational imaging
• Segmentation, grouping and shape representation
• Object detection, recognition and re-identification
• Document analysis, handwriting and text recognition

Video Analysis and Understanding

• Video classification and representation
• Action recognition, event understanding and video segmentation
• Object tracking and motion analysis
• Human behaviour understanding, pose and gesture analysis
• Egocentric (first-person) vision

Multimedia + Vision and Language

• Vision–language models, multimodal learning and visual reasoning
• Multimedia indexing, retrieval and content analysis
• Cross-modal understanding and generation

3D Computer Vision and Geometry

• 3D reconstruction, structure from motion and SLAM; visual localization
• Neural rendering and neural scene representations
• Geometry processing, shape matching and point-cloud analysis
• Augmented, virtual and mixed reality

Medical Imaging and Health Technologies

• Medical and biomedical image analysis
• Surgical video understanding
• Biological vision and microscopy
• Assistive technologies and health monitoring

Digital Forensics and Biometrics

• Biometrics: face, fingerprint, iris and other modalities
• Digital forensics, media authentication and synthetic-media detection
• Adversarial attacks and defenses; security and privacy in vision

Cultural Heritage and Visual Creativity

• Image analysis for cultural heritage and digital humanities
• Digitization, restoration and analysis of artworks and historical documents
• Visual creativity, computational art and design

Robot and Vision

• Robot vision and visual perception for navigation
• Embodied vision: active agents and simulation
• UAV imagery and autonomous driving
• Synthetic data and simulation

AI and Science

• Vision and AI for the physical, life and environmental sciences
• Remote sensing and Earth observation
• Climate and sustainability

Applications

• Vision applications and systems in industry, agriculture, retail and other domains
• Efficient, embedded and edge vision
• Social-good applications

Submission guidelines

Detailed instructions will accompany the opening of the submission system. Authors should note the following requirements:

• Format — Springer LNCS style, single-column, using the official ICIAP 2027 Author Kit (LaTeX).
• Length — at most 12 pages, including figures, tables and references.
• Anonymity — submissions are reviewed double-blind and must not identify the authors; see the Author Kit for details.
• Submission system — OpenReview; the link to the submission page will be published here when submissions open. All authors must have an OpenReview profile and will be asked to provide a preferred email address.
• Rounds and deadlines — submissions are accepted in two rounds (see Important dates). There is no separate abstract registration: papers are submitted in a single step, at the paper submission deadline. Deadlines are at 23:59 AoE (Anywhere on Earth).
• Supplementary material and code — no supplementary material is foreseen. Authors may, however, share the source code and the instructions to reproduce the experiments through the dedicated fields of the submission form, either as a link to an anonymised repository on anonymous.4open.science (preferred) or as an anonymised zip file. During review the code must be anonymous, like the paper, and the paper itself must not contain links to the code. Public releases (for example on GitHub) are expected only after acceptance; see the RRPR Badge.
• Language — all submissions must be written in English.
• Public metadata — the title, abstract and author list of accepted papers will be made publicly available on OpenReview.

Review process

Following the practice of recent editions, ICIAP 2027 runs submissions in two rounds, so that authors have a second opportunity to submit new work or revised manuscripts. Papers rejected in the first round may be resubmitted to the second round; in that case, we will do our best to assign them to the same reviewers. Each submission is peer reviewed by members of the program committee; the program chairs make the final decision on acceptance and on the assignment of papers to oral or poster sessions.

Reproducibility: the RRPR Badge

ICIAP 2027 will award the Reproducible Research in Pattern Recognition (RRPR) Badge, promoted by IAPR Technical Committee 22, to accepted papers whose code allows their results to be reproduced. The process follows the one adopted at ICPR 2026 (see the RRPR Badge guidelines). Applying is optional and has no effect on the scientific review of the paper.

• How to apply — tick Apply for the RRPR Badge in the OpenReview submission form.
• Code at submission — the request is considered only if the code is provided at the time of submission, through the code fields of the submission form. An anonymised repository on anonymous.4open.science is preferred over the zip file, as the same repository can then be made public as soon as the paper is accepted. Applications without code at submission time are not considered.
• Accepted papers only — the paper must first be accepted through the regular review process. Only accepted papers are evaluated for reproducibility.
• Public code before the camera-ready — to remain eligible, authors must make the code publicly available immediately after acceptance, and in any case before the camera-ready deadline. The public repository must explain how to install the dependencies and run the code, and the data needed to reproduce the results must be publicly available.
• Evaluation — the released code is checked by reviewers coordinated by the Reproducible Research Chairs, following the RRPR Badge guidelines.

Policies

• Dual submission — by submitting a paper, authors confirm that no substantially similar paper has been or will be submitted to another conference, workshop or journal during the review period, that is, from the submission deadline to the notification date of the round in which the paper is submitted.
• Preprints — posting a preprint on a public server such as arXiv is allowed. Submissions must nevertheless remain anonymous: authors must not cite or link their own preprint in a way that identifies them.
• Ethics and plagiarism — authors must follow the Springer Nature Code of Conduct for Authors. Submissions may be checked for plagiarism; plagiarism, fabricated or falsified results and unacknowledged reuse of previously published material are grounds for rejection. Work involving human subjects or personal data must comply with the applicable ethical approval and informed-consent requirements.
• Conflicts of interest — conflicts are computed by OpenReview from the information in the authors’ profiles (affiliations, co-authors and relations), so authors should keep their profile up to date. Reviewers and area chairs are not assigned papers with which they have a conflict.
• Reviewer service — for each submitted paper, at least one of its authors will be required to serve as a reviewer and will be assigned 2–3 papers to review. The load applies per submitted paper: an author designated as reviewer for several submissions will be assigned 2–3 papers for each of them. Failure to complete the assigned reviews may lead to the desk rejection of the submission.
• Use of LLMs and generative AI — the use of large language models and other generative AI tools is permitted. Authors who have used these tools must disclose how, in a dedicated section included within the 12-page limit.
• Disclosure of Interests — a short Disclosure of Interests statement is mandatory in the camera-ready version of accepted papers, as required by Springer LNCS.

Publication

Accepted papers will be published by Springer in the Lecture Notes in Computer Science (LNCS) series.

Workshops and tutorials

Separate calls will be issued for workshop and tutorial proposals. These events take place alongside the main conference. Proposers should watch the program page for the corresponding deadlines.

Contact

For questions about this call, please write to iciap2027-info@unimore.it.