eccv

name

eccv

description

Writing and formatting guidance for ECCV (European Conference on Computer Vision) papers. Use when drafting, formatting, and submitting to ECCV 2026. ECCV uses a Springer LNCS base with the eccv.sty overlay. Covers eccv document class with year option, Springer LNCS formatting conventions (llncs.cls), double-blind anonymity with 14-page limit, Springer-style bibliography (splncs04), and camera-ready preparation.

ECCV 2026 Conference Writing Skill

CCF-B | AI | Publisher: Springer Conference: https://eccv.ecva.net/Conferences/2026 Template: ccf-latex-templates/ECCV/main.tex Base: Springer LNCS (llncs.cls)

Document Setup

Preamble Structure

\documentclass[runningheads]{llncs}

% === MODE SELECTION ===
% Review version:
\usepackage[review,year=2026,ID=*****]{eccv}
% OR for camera-ready:
% \usepackage{eccv}

% Abbreviations:
\usepackage{eccvabbrv}

% Packages:
\usepackage{graphicx}
\usepackage{booktabs}
\usepackage{amsmath}
\usepackage{amssymb}
\usepackage{subcaption}

% Hyperref:
\usepackage[pagebackref,breaklinks,colorlinks,citecolor=eccvblue]{hyperref}

Mode Options

Mode	Declaration	Features
Review	\usepackage[review,year=2026,ID=*****]{eccv}	Line numbers, anonymized, paper ID on each page
Camera-Ready	\usepackage{eccv}	No line numbers, author info

Replace ***** with your assigned paper ID.

Page Limits

• Main content: 14 pages maximum
• Excluded: References, acknowledgments
• Papers exceeding 14 pages will be rejected without review
• This is a generous limit compared to CVPR/ICCV

Anonymity Requirements (Double-Blind)

1. No author identification in submission
2. Use \author{} with real names only for camera-ready
3. Third-person self-citations ("Smith et al. showed..." not "we showed...")
4. Remove acknowledgments from submission
5. Anonymize identifying URLs
6. If citing concurrent submission: cite anonymously, include anonymized copy as supplementary

Title and Author Formatting

\title{Full Paper Title}

% Short title for running head:
\titlerunning{Abbreviated paper title}

% Full author list (for camera-ready):
\author{First Author\inst{1}\orcidlink{0000-1111-2222-3333} \and
Second Author\inst{2,3}\orcidlink{1111-2222-3333-4444} \and
Third Author\inst{3}\orcidlink{2222-3333-4444-5555}}

% Abbreviated for running head:
\authorrunning{F. Author et al.}

% Full affiliations:
\institute{
  Princeton University, Princeton NJ 08544, USA \and
  ABC Institute, Rupert-Karls-University Heidelberg, Germany\\
  \email{\{abc\}@uni-heidelberg.de}
}

Abstract and Keywords

\begin{abstract}
The abstract should concisely summarize the contents of the paper.
While there is no fixed length restriction, approximately 150 words is recommended.
Please include keywords as shown below.
\keywords{First keyword \and Second keyword \and Third keyword}
\end.Abstract}

Recommended CV Paper Structure for ECCV

\section{Introduction}
% Paragraph 1-2: Problem and motivation
% What is the task? Why is it important?

% Paragraph 3: Gap in prior work
% What have others done? What are their limitations?

% Paragraph 4: Our approach
% What is our key insight?

% Paragraph 5: Contributions (numbered)
The main contributions of this work are:
\begin{itemize}
    \item We propose a novel approach to X that achieves Y...
    \item We introduce a new benchmark dataset for...
    \item We demonstrate through extensive experiments that...
\end{itemize}

\section{Related Work}
% Categorize prior work by approach type
% Position your work against each category

\section{Preliminaries}
% Notation, definitions, background concepts

\section{Method}
% Core technical contribution
% Include figures showing architecture

\section{Experiments}
\subsection{Setup}
% Datasets, metrics, baselines, implementation

\subsection{Results}
% Main findings in tables
% Comparison with state-of-the-art

\subsection{Ablation and Analysis}
% Study each component
% Qualitative results

\section{Conclusion}
% Summary, limitations, future work

Figures and Tables

Figure Guidelines

\begin{figure}[tb]
  \centering
  \includegraphics[height=6.5cm]{figure}
  \caption{
    One kernel at $x_s$ (\emph{dotted kernel}) or two kernels at $x_i$ and $x_j$.
    Elements described in caption should be set in italics, in parentheses.
  }
  \label{fig:example}
\end{figure}

% Full-width figure:
\begin{figure*}[tb]
  \centering
  \includegraphics[width=0.9\textwidth]{figure}
  \caption{Full-width figure caption.}
  \label{fig:wide}
\end{figure*}

• Caption below figures
• Use vector formats (PDF, EPS)
• 800 DPI minimum for line drawings
• 6pt minimum font size in figures

Table Guidelines

\begin{table}[tb]
  \caption{
    Font sizes of headings.
    Table captions should always be positioned \emph{above} the tables.
  }
  \label{tab:headings}
  \centering
  \begin{tabular}{@{}lll@{}}
    \toprule
    Heading level & Example & Font size\\
    \midrule
    Title & 14pt, bold & centered \\
    1st-level & 12pt, bold & left \\
    \bottomrule
  \end{tabular}
\end{table}

• Caption above tables
• Use booktabs
• No vertical rules

Vision-Specific Writing Conventions

Dataset Description

\section{Experiments}
\subsection{Datasets}
We evaluate on standard vision benchmarks:

\begin{itemize}
    \item \textbf{COCO}: 118K training, 5K validation images.
    \item \textbf{ADE20K}: 20K training, 2K validation images.
    \item \textbf{Cityscapes}: 2,975 training, 500 validation images.
\end{itemize}

Evaluation Metrics

\subsection{Metrics}
We report standard vision metrics:
\begin{itemize}
    \item \textbf{Object detection:} mAP@0.5:0.95 (COCO standard)
    \item \textbf{Semantic segmentation:} mIoU
    \item \textbf{Instance segmentation:} AP$^{bbox}$, AP$^{mask}$
    \item \textbf{Panoptic segmentation:} PQ
\end{itemize}

References (Springer LNCS Style)

\bibliographystyle{splncs04}
\bibliography{main}

Citation format:

\cite{Author22}           % [1]
\citep{Author22}          % [1]
\citet{Author22}          % Author (2022)

Heading Styles (LNCS)

Level	Format	Numbering
Title	14pt, bold, centered	No
1st-level	12pt, bold, left	Yes
2nd-level	10pt, bold, left	Yes
3rd-level	10pt, bold, left	No

Capitalize headings (first word + content words).

Camera-Ready Compilation

1. Comment out review mode:

% \usepackage[review,year=2026,ID=*****]{eccv}
\usepackage{eccv}

2. Add complete author information with ORCID
3. Include acknowledgments (unnumbered section)
4. Remove line numbers
5. Update hyperref settings
6. Verify 14-page limit

Checklist Before Submission

• 14 pages or fewer (excluding references)
• Paper ID on each page (review mode)
• No author identification
• Third-person self-citations
• References in splncs04 format
• Figure captions below, table captions above
• Heading capitalization correct
• No \vspace or \hspace modifications
• DOIs included in references
• Acknowledgments removed from submission