ACM Transactions on Graphics Submission Guide

Source: ACM TOG on ACM DL, SIGGRAPH Asia 2026 Technical Papers, Clarivate JCR 2024, accessed April 2026.

This is the ACM TOG submission feature most authors don't know:

SIGGRAPH and SIGGRAPH Asia conference papers are simultaneously ACM Transactions on Graphics journal papers.

Authors who submit to:

• SIGGRAPH (typically annual deadline in January), papers appear in the SIGGRAPH proceedings AND in TOG

• SIGGRAPH Asia (typically annual deadline in May), papers appear in the SIGGRAPH Asia proceedings AND in TOG

The journal-track integration means most TOG papers go through one of the two conference cycles per year. Direct TOG submissions outside these cycles are accepted but less common; most authors prefer the conference path because it provides community visibility and a structured timeline.

The strategic implication: pick the conference path that matches your timeline. Submitting to SIGGRAPH Asia 2026 (May 2026 deadline) means publication later in 2026; submitting to SIGGRAPH 2026 (January 2026 deadline already passed) means waiting for SIGGRAPH 2027 if not making it through SIGGRAPH Asia. Direct TOG submission outside the conference cycle is the alternative for time-sensitive work.

Page limits depend on the submission path:

The strategic implication: SIGGRAPH conference-track papers force compression to a 7-page core plus figure-only pages, while journal-only and direct TOG submissions can be longer if the contribution needs that space. Authors who present preliminary work at SIGGRAPH and develop the full method paper at TOG outside the conference cycle is one route, but most graphics work goes through the conference cycle.

Verify the current Editor-in-Chief on the journal's editorial-team page before quoting any name in a cover letter. The Editor-in-Chief oversees the editorial board that handles direct TOG submissions and coordinates with SIGGRAPH and SIGGRAPH Asia paper chairs for the journal-track conference papers.

The practical consequence: editorial direction reflects the broader graphics-research community. Recent TOG issues span the full breadth of computer graphics:

• Neural rendering and learned graphics methods

• Differentiable rendering and inverse graphics

• Geometry processing and shape analysis

• Animation, simulation, and physics-based modeling

• Computational displays and computational photography for graphics

• Real-time rendering and ray tracing

• Generative models for 3D content

This is what editors check before review: whether the contribution is genuinely graphics methodology and whether the visual evidence package proves it.

ACM TOG's editorial filter (and SIGGRAPH/Asia paper-chair filter) turns on three operational signals:

1. The contribution is graphics-specific methodology. TOG publishes graphics methodology. A paper that uses graphics techniques as a tool for non-graphics research (e.g., image classification using rendered images) often fits CS venues like CVPR or NeurIPS better. The first-paragraph test: would a graphics methodologist find the methodological contribution useful, not just the application?

2. Comparison to state-of-the-art on standardized graphics benchmarks. TOG and SIGGRAPH expect comparison to recent graphics methods using established benchmarks (e.g., Stanford 3D Scanning Repository for geometry, NeRF / 3D Gaussian Splatting baselines for neural rendering, established physics simulation benchmarks). Manuscripts with limited baseline comparison face higher review scrutiny.

3. Visual quality and artistic merit are real evaluation criteria. TOG papers are evaluated partly on the quality of the graphics produced. Papers with thin visual results, low-quality figures, or limited demonstrations face additional scrutiny that wouldn't apply at theory-only CS journals. Investing in high-quality figures, supplementary videos, and visual comparisons is part of the submission package.

Before submitting to ACM Transactions on Graphics, an ACM Transactions on Graphics submission readiness check identifies whether the package meets the editorial bar before you commit to the submission.

Recent TOG / SIGGRAPH issues span:

• Neural rendering (NeRF, 3D Gaussian Splatting, neural fields for graphics)

• Differentiable rendering and inverse graphics methods

• Generative models for 3D content (diffusion models, GANs for graphics)

• Geometry processing for meshes and point clouds

• Real-time rendering and ray tracing on modern hardware

• Animation, character rigging, and physics simulation

• Computational displays, AR/VR rendering, perception

For specific recent papers and DOIs, see the ACM TOG archive. The DOI prefix is 10.1145/* with paper-specific identifiers.

Recent TOG and SIGGRAPH-track examples show the range of accepted graphics evidence: differentiable geometric acoustic path tracing (10.1145/3730900), moment-bound approximation for inverse rendering (10.1145/3730899), and conforming weighted Delaunay triangulations for fixed-dimension geometry processing (10.1145/3763368). Use recent examples to calibrate visual proof, benchmark choice, and supplemental evidence, not to copy topical framing.

Submission caps: SIGGRAPH 2026 dual-track conference papers cap the core paper at 7 pages excluding references, plus up to 2 figures-only pages. Journal-only SIGGRAPH 2026 submissions and direct ACM Transactions on Graphics submissions do not have a formal maximum or minimum page length, but the paper still needs to be concise enough for reviewers to follow the graphics contribution. ACM expects supplemental materials to be submitted electronically; videos are strongly encouraged in MP4, and supplemental packages above 500 MB are not guaranteed to be downloaded and viewed.

For SIGGRAPH or SIGGRAPH Asia conference paper submission:

1. Manuscript within the conference page limit (7 core pages excluding references + up to 2 figures-only pages for SIGGRAPH 2026 dual-track submissions)

1. Supplementary material for additional results, experiments, derivations

1. Supplementary video demonstrating graphics results (often expected for animation, simulation, rendering papers)

1. Title page, authors, affiliations (anonymized for double-blind review at the conferences)

1. Cover letter or summary describing the contribution

1. Code or model release plans (community expectation though not always enforced)

1. Reviewer suggestions as the conference allows

For direct TOG submissions outside conferences:

1. Manuscript in standard ACM Transactions journal format

1. All supporting materials as above

1. Cover letter explaining why a direct TOG submission rather than a conference path

1. ORCID identifiers for all authors (required for ACM submissions)

1. Author contributions statement following ACM CRediT taxonomy

1. Conflicts of interest disclosure for all authors

1. Funding statement disclosing grants, sponsor support, or institutional funding

1. Ethics statement where human-subjects studies, user evaluations, or sensitive data are involved

1. Data and code availability statement; ACM increasingly requires artifact links

1. Supplementary video demonstrating the visual contribution (required for animation, simulation, rendering papers)

A ACM TOG submission readiness check before submission can flag whether the contribution is graphics-specific (vs vision or ML), whether the page count fits the chosen path, and whether supplementary video/visual quality meets community expectations.

For TOG, the pre-submission check should include the paper and the visual evidence package. A graphics paper can read strong in prose and still fail if the supplementary video, comparisons, or figure sequence do not make the method's visual advantage obvious.

Before upload, an ACM Transactions on Graphics readiness check can catch the difference between a visually convincing TOG package and a paper that belongs at CVPR, CHI, NeurIPS, or a narrower graphics workshop.

Run the free manuscript readiness scan before you submit if you want the graphics-specific contribution, page limit, and visual-evidence package checked together.

TOG's flow follows the SIGGRAPH conference cycle for journal-track submissions and the ACM Transactions process for direct submissions. Treat as planning ranges, not promises.

• Day 0: ACM submission or conference deadline upload. SIGGRAPH Asia 2026 deadline lands in May; SIGGRAPH 2026 deadline in January. Direct TOG submissions are rolling at ScholarOne submission portal.

• Days 1 to 14: Editorial admin and paper-chair assignment. Editors (or SIGGRAPH paper chairs for the conference track) verify page-limit compliance, anonymization for double-blind review, and supplementary video requirements.

• Days 14 to 90: Peer review. Three or more reviewers evaluate the graphics methodology, visual evidence, comparison-baseline fairness, and reproducibility. Conference papers and direct TOG submissions follow the same review depth.

• Days 90 to 120: Author response and reviewer discussion. Conference cycle includes a formal author-response window; direct TOG follows the standard ACM rebuttal process.

• Days 120 to 150: First decision. Major revision is the most common outcome for papers that pass review. Conditional accept is also common at the conferences.

• Days 150 to 240: Revisions, final decision, and publication. Conference papers appear at the SIGGRAPH conference plus the corresponding TOG journal issue; direct TOG papers appear in a regular TOG issue.

Read the public instructions for mechanics, then pressure-test the package the way an editor will see it. The review tells you whether your paper clears the ACM Transactions on Graphics fit check before upload, especially around three failure modes: applied ML that is not graphics methodology, SIGGRAPH page-count compression, and visual evidence below the community bar.

Paid Manusights reviews include a 60-day money-back guarantee, and we do not train models on submitted manuscripts.

Across graphics manuscripts targeting ACM Transactions on Graphics, three recurring decision risks matter most across submissions that TOG editors and SIGGRAPH conference-track reviewers filter out at the desk-screen or first-review stage. (Per SIGGRAPH 2026 Technical Papers FAQ, three or more reviewers evaluate graphics methodology, visual evidence, comparison-baseline fairness, and reproducibility;

failure to follow the anonymous-cover-letter scope-justification policy triggers desk rejection without review, as do dual submissions, incomplete writeups, advertising-shaped papers, off-scope topics, and rights violations.) Use the three checks below before you open SIGGRAPH submission system or the direct TOG ScholarOne upload.

A Manusights review checks whether your paper clears the ACM TOG-specific readiness checks that official ACM and SIGGRAPH instructions cannot evaluate from a generic submission-system checklist. Paid Manusights reviews include a 60-day money-back guarantee; submitted manuscripts are not used for model training.

Applied ML without graphics-method novelty

Across TOG-targeted manuscripts, we consistently see authors submit work where the contribution is computer-vision methodology (object detection, segmentation, depth estimation, action recognition, 3D reconstruction from photos), machine-learning methodology (new architecture, new loss, new training scheme, new self-supervised objective), or natural-language methodology (text-to-3D, multimodal generation) applied to graphics-relevant data, but the methodological advance is in the CV / ML / NLP technique rather than in graphics.

TOG and SIGGRAPH reviewers (drawn from the graphics community) specifically check whether the contribution is graphics methodology (rendering algorithm, geometry processing operator, animation control, physical simulation method, geometric optimization, light-transport advance, display-system contribution, character-animation primitive, fluid / cloth / hair / collision-detection method) or whether graphics is the application domain for an underlying CV / ML / NLP technique.

Manuscripts where the contribution is non-graphics methodology applied to graphics data face desk rejection or early-cycle rejection, with reviewer comments pointing to CVPR / ICCV / ECCV / NeurIPS / ICLR / ICML / EMNLP / 3DV / ECCV-Workshops / SIGGRAPH-Workshops as better venues.

The fix is to either (a) reframe the contribution so the graphics methodology is load-bearing (e.g., "we present a new differentiable rendering formulation that, as a downstream consequence, enables faster novel-view synthesis" rather than "we present a NeRF variant"), and demote the ML architecture to a methodological tool; or (b) route honestly to the CV / ML venue where the contribution actually lives.

The intermediate case (graphics methodology that uses ML as a tool) fits TOG; the inverse (ML methodology applied to graphics) does not.

Check whether your ACM TOG manuscript is graphics methodology rather than applied ML →

SIGGRAPH page-count compression

In Manusights reviews, we observe that authors with prior NeurIPS / ICLR / ICML / CVPR / ECCV submission experience routinely overshoot SIGGRAPH's 7-page dual-track conference-paper core, excluding references and figures-only pages.

The submission system enforces a hard page-count check, but more important is the editorial expectation: papers that arrive at exactly 7 pages with the methodology, results, comparisons, and discussion all crammed in (rather than developed) get review comments saying the paper feels rushed.

SIGGRAPH's reviewer-comment culture specifically calls out: methodology sections that skip derivations or design rationale, comparison sections that show fewer than the expected 3-5 baselines per task, ablation sections that omit obvious component-removal experiments, results sections without per-scene / per-input quantitative metrics in the main paper (not just supplementary), and discussion sections that fail to address failure modes and limitations.

The fix is to:

• structure the paper for the 7-page cap from the first draft (not by compressing a longer draft)

• move algorithm pseudocode + extended derivations + additional baselines + additional ablations + per-scene metric tables to supplementary.

• use the figures-only pages for high-density comparison grids and ablation montages rather than as overflow text.

• ensure the supplementary video is the load-bearing visual artifact (3-5 minutes, narrated, showing the method on diverse inputs with side-by-side baseline comparisons)

• treat the 7-page main paper as the methodological argument with supplementary carrying the evidence.

Check whether your ACM TOG page-count and supplementary strategy fit the SIGGRAPH path →

Visual evidence below the SIGGRAPH bar

The third recurring pattern in TOG-targeted manuscripts is a visual-evidence package below the SIGGRAPH community bar.

SIGGRAPH and TOG reviewers evaluate methodological contribution partly through visual quality and reproducibility. Common evidence gaps include:

• low-resolution figures, below 300 DPI for raster images or below vector quality for diagrams

• figures without anti-aliasing

• comparison rows where the proposed method's image is larger or differently cropped than baselines

• missing supplementary video for motion, animation, simulation, interaction, time-varying, or real-time-performance claims

• supplementary video without baseline comparisons, or with cherry-picked single examples

• no source code release or executable demo where the community expectation exists

• no per-pixel, per-frame, or per-scene metric tables alongside averaged headline numbers

• no failure-case section showing where the method breaks

Reviewers also check that comparison-baseline fairness is evident: baselines are run with their authors' published code and parameters, not authors' reimplementations; baselines are tested on the same inputs as the proposed method; metric definitions match published baselines exactly.

The fix is to invest in high-DPI figures with consistent comparison crops, prepare a 3-5 minute narrated supplementary video with side-by-side baselines (not just method-only), release source code or an executable demo before submission (anonymized via a temporary repository where required), include per-input quantitative tables in supplementary, and add an explicit failure-cases section showing where the method breaks.

Check whether your ACM TOG visual evidence package is strong enough →

• Is ACM Transactions on Graphics a good journal?

• Communications of the ACM submission guide

• International Journal of Computer Vision submission guide

• SIGGRAPH Submission Guide

• IEEE TPAMI Submission Guide