Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› siggraph-topic-selection
siggraph-topic-selection
GitHub辅助决策SIGGRAPH系列会议及TOG期刊的投稿策略。通过视觉结果、能力轴和方法论测试判断贡献是否属于图形学范畴，并根据技术类型路由至SIGGRAPH主会、Asia、TOG或CVPR/CHI等兄弟会议，同时提供周期选择建议。
触发场景
准备向SIGGRAPH/TOG投稿前评估选题
不确定项目是图形学还是计算机视觉/HCI贡献
需要在多个相关会议（如CVPR, CHI, EGSR）中选择合适的投稿目标
安装
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill siggraph-topic-selection -g -y
SKILL.md
Frontmatter
{
"name": "siggraph-topic-selection",
"description": "Use before committing a project to SIGGRAPH \/ SIGGRAPH Asia \/ TOG, to decide whether the contribution is graphics-shaped and which venue and track fit, routing among SIGGRAPH, SIGGRAPH Asia, direct-to-TOG, and siblings (CVPR\/ICCV, CHI\/UIST, Eurographics\/EGSR\/SGP\/SCA\/HPG\/I3D), and whether to aim dual-track or Journal-only."
}
SIGGRAPH Topic Selection
Choosing SIGGRAPH — and the right cycle and track — before you write is a high-leverage decision,
because SIGGRAPH's bar is a graphics contribution shown through results, and several sibling
venues will serve a differently-shaped contribution better. This skill routes the project. Anchor
venue facts to resources/official-source-map.md.
Is it a graphics contribution?
SIGGRAPH rewards work that advances how images, geometry, motion, or interactive experiences are created, simulated, rendered, captured, or fabricated — and demonstrates it on results a graphics audience can see. Apply three tests:
- The visual/temporal-result test. Can the contribution be shown as an image, mesh, frame sequence, or interactive demo that a graphics reviewer judges by eye? If the payoff is a table of accuracy numbers with nothing to show, it is likely a vision or ML paper.
- The capability-axis test. Does it move a graphics axis — quality, speed, generality, robustness, controllability? If there is no such axis, the contribution is not yet SIGGRAPH-shaped.
- The method-vs-application test. Is the advance a reusable technique, or an application of an existing method to a new dataset? The latter routes to a domain venue or a shorter track.
Route the project
| If the contribution is... | Consider | Why not SIGGRAPH main |
|---|---|---|
| A graphics technique with strong visual results | SIGGRAPH / SIGGRAPH Asia Technical Papers (TOG) | This is the target |
| Rendering-specific, incremental but solid | EGSR or HPG (and TOG) | Better-matched reviewers; faster cycle |
| Geometry-processing-specific | SGP (and TOG/CGF) | Specialized geometry audience |
| Animation / physical simulation focused | SCA (and TOG) | Sub-community venue |
| Interactive/real-time rendering technique | I3D or HPG | Real-time-focused reviewers |
| Recognition/reconstruction where the model is the point | CVPR / ICCV / ECCV | A vision contribution, not a graphics one |
| Interaction/UI/hardware where usability is the point | CHI / UIST | An HCI contribution |
| A mature, comprehensive result with no conference talk needed | Direct-to-TOG (rolling) | Journal-only, off-cycle |
| Early-stage but promising graphics idea | SIGGRAPH Conference track (dual-track) or Technical Communications | Lighter validation bar |
The neural-rendering / generative-3D overlap with CVPR is the most common routing decision: if the lesson survives swapping the network (a graphics capability), aim SIGGRAPH; if the network is the contribution and the graphics is a testbed, aim a vision venue.
Choose the cycle
- SIGGRAPH (North America), ~January deadline, summer conference vs SIGGRAPH Asia, ~May deadline, December conference. Same TOG journal outcome, same review model — pick by which deadline your results can finish to the journal bar, and by conference travel.
- A near-miss at one cycle is naturally revised for the next; do not rush an unfinished paper into the nearer deadline.
- Direct-to-TOG (rolling, ScholarOne) suits a comprehensive journal paper that does not need a conference talk; most authors still go through the conference cycle for the presentation and visibility.
Choose the track
- Dual-track (default): eligible for Journal (TOG) or Conference publication; keeps the
lighter-validation Conference outcome available for riskier/earlier-stage work. The committee
assigns the final track — you cannot appeal it (see
siggraph-review-process). - Journal-only: choose when a Conference-track publication would not serve a mature, complete result. This forgoes the lighter-validation safety net.
- You do not choose between journal and conference reviewers — there is one review on one scale; the track is decided at accept time.
Honest self-check before committing
- Do I have a results video and head-to-head comparisons against the strongest prior method, or can I produce them by the deadline? Without them, SIGGRAPH is a long shot.
- Is the closest prior work at SIGGRAPH, or at Eurographics/EGSR/SGP/SCA/CVPR? Submit where your reviewers and canon live.
- Is the contribution a reusable technique or a one-off result? The former is SIGGRAPH-shaped.
Output format
[Graphics-shaped?] visual-result / capability-axis / method-vs-application tests: pass/fail
[Venue] SIGGRAPH / SIGGRAPH Asia / TOG-direct / sibling (name it) + why
[Cycle] which deadline the results can credibly finish to
[Track] dual-track / journal-only + rationale
[Readiness] results video + strongest-baseline comparison in hand or schedulable? yes/no
[Route decision] <one line + next skill to invoke>
版本历史
- 9f86f09 当前 2026-07-19 17:36
