维护入口汇总：index.html 管正文与版面；data/site-config.json 管全站资源与发布信息；data/samples.json 管样本索引。本页 HTML 与 Markdown 应同步维护。
PACTHub content map and materials checklist
Data standardization and the image-enhancement algorithm are already complete. This checklist organizes only the materials needed for website presentation, downloads, and reproducibility; it does not imply that the research still needs to be carried out. Names, scale, results, people, and links that have not been supplied remain pending. Illustrative content must not stand in for research conclusions.
Recommended reading order
Project overview: What is this, who is it for, and what resources are available?
Samples and previews: What does the data look like, and which fields affect its use?
Image-enhancement algorithm: What does the method do, what are its inputs and outputs, and what supports its results?
Paper and citation: Where is the research evidence, and how should I cite it?
Team and contributions: Who maintains this, and where should questions or contributions go?
Help visitors understand the resources first, assess suitability next, and then start using them. Keep core information on the homepage and link to the existing guides for full data specifications and algorithm steps.
01 · Project overview
Question to answer: Which research tasks does PACTHub support, and which completed outputs does it provide?
Required for public launch
Text: Confirm the public project name, a one-sentence positioning statement, and a short introduction of roughly 100–180 Chinese characters or equivalent English length, explaining the study subject, use cases, and relationship between data and algorithm.
Facts: Provide the data version and coverage approved for public release, and explain how counts were calculated. Do not display unverified totals, performance, or coverage.
Hero video/poster: Provide an approved public video and matching static poster, with source and intended-use notes. The project-provided 20-second video and matching poster are now included. Confirm publication rights and descriptions before public deployment.
Image: Optionally include one representative image approved for publication, with source, caption, and alternative text. Omit it when no real image is available; do not present a scientific illustration as measured results.
Links/files: Establish at least one primary action, such as browsing samples, reading instructions, or obtaining data. No separate overview download is needed.
Possible later additions: Project logo, more bilingual project background, and additional application examples.
Edit here:site_title, description, and overview_summary in data/site-config.json; index.html#overview controls overview structure and imagery. release_version stores the release record and currently has no separate display. Add only verified numbers to the page.
02 · Samples and previews
Question to answer: What data can I inspect, and what do its spatial scale, provenance, and groups mean?
Required for public launch
Text: A unique ID, title, short description, real category, and actual dataset split for each public sample. Explicitly mark absent splits as unspecified; do not reuse DEMO groups.
Fields: Supply publicly shareable dimensions, axis order, voxel spacing, units, wavelength, and provenance under the existing data standard, including standard and data versions. Represent unknown or inapplicable fields according to the current Schema; do not guess values.
Images: Start with 3–6 representative real samples if possible, with at least one for validating presentation. Each preview needs a caption describing slice/projection method, orientation, scale, and display processing. State clearly when images cannot be public.
Links/files: Include corresponding preview images and each released file's name, format, download URL, actual byte count, and SHA-256. For controlled data, link only to an approved access-request route.
Possible later additions: More categories, three-view or slice browsing, filter tags, and lightweight 3D previews. New interactions require separate implementation; they are not existing capabilities of the static site.
Edit here:index.html#explore is the browsing entry point. Store sample records in data/samples.json, and previews in assets/images/ or at verified HTTPS URLs. See the data integration guide for field mapping.
03 · Image-enhancement algorithm
Question to answer: What does the completed algorithm improve, how is it run, and when is it unsuitable?
Required for public launch
Text: Official algorithm name; a short method summary of roughly 100–180 Chinese characters or equivalent English length; inputs and outputs; relationship to the existing data standard; scope and known limitations.
Images: If showing results, supply at least one real input/output comparison from the same sample, with captions. Keep orientation, crop, and display scale comparable, and explain processing. Make no performance claims without comparison images approved for publication.
Evidence: When showing metrics, include their definitions, test data/splits, reference ground truth, execution conditions, and result sources. Omit performance numbers without verified results.
Links/files: A code or method-documentation entry point and specific release version. If a model is released, provide its matching weights and license. Prepare dependency/environment files, configuration, and a verified minimal inference example. Explicitly mark inapplicable materials.
Possible later additions: Method flowchart, more results, ablation studies, and runtime comparisons. Online inference requires a separately developed service.
Edit here: Structure and result images in index.html#algorithm; algorithm.name, summary, status, repository_url, and weights_url in data/site-config.json provide text and resources. Maintain full steps in the algorithm integration guide.
04 · Downloads and usage
Question to answer: Which package fits my needs, how do I verify, read, and reproduce it, and what use restrictions apply?
Required for public launch
Text: Actual file formats and directory structure, minimal reading steps, verified environment and commands, data license, code/weight licenses, and access restrictions.
Files: At least one verifiable lightweight sample and reading example, or a clear access-request process. Record version, size, and SHA-256 for each released package, with a README. Do not publish empty packages or placeholder download links.
Links: Real entry points for the dataset, lightweight sample, and algorithm resource package. Keep unavailable items marked pending or access-on-request. Confirm external repository permissions and link validity.
Images: Optional. Add diagrams only when text cannot explain the file structure or reading process clearly.
Possible later additions: Notebooks, containers, frequently asked questions, release notes, and reading examples in more languages.
Edit here:downloads and license in data/site-config.json; usage summary in index.html#downloads. Add reading and reproducibility details to the data and algorithm guides respectively. List individual sample files in each record's files in data/samples.json.
05 · Paper and citation
Question to answer: Which paper supports the results, and which resource version should I cite?
Required for public launch
Text: Exact title, author order, year, publication status, and venue/source for the public paper; citation text or BibTeX matching the actual resource version.
Links/files: An existing, verified DOI, paper, or preprint URL. Offer PDFs only with distribution rights.
If no paper can be public yet: Explain its current availability and give a project-approved way to cite the resources. Keep the DOI empty if none exists; do not construct a URL.
Images: Optional; paper screenshots are not needed to fill this section.
Possible later additions: A separate .bib file, instructions for citing the dataset and algorithm separately, and related papers.
Edit here:paper.title, paper.summary, paper.url, citation, and doi in data/site-config.json; index.html#publication is the corresponding page section. A nonempty citation enables copying. Enter a real, complete 10.digits/suffix identifier in doi; valid syntax generates a DOI link.
06 · Team and contributions
Question to answer: Who maintains this, and how do I report an issue or contribute?
Required for public launch
Text: Confirm publicly shareable maintainer/team names and affiliations, maintenance scope, and short contribution and feedback instructions.
Links: At least one usable public contact route, such as repository Issues or a team contact page. An approved public email may appear in explanatory text. Distinguish routes for technical issues, data requests, and collaboration.
Images/files: Optional. Confirm permission to publish names, portraits, affiliation logos, and acknowledgments.
Possible later additions: Member profiles, portraits, funders/partner institutions, full contribution guidelines, and a contributor list.
Edit here:team.members (each item has name, affiliation, role, and url), team.contribution, and team.contribution_url in data/site-config.json; structure in index.html#team. Links accept only HTTP(S); contact entry points do not support mailto:. Do not include personal contact details without permission.
Minimum materials package to supply together
Keep original image files and provide publishable captions. Host archives and model weights in an approved data repository rather than putting them into the website repository. Materials may follow the existing standard; the website maps them and does not require a new data standard.
Configuration and launch checks
Keep unknown information as null and absent members as []. Empty configurable text retains the HTML default explanation; missing resources do not enable links. New configuration fields require corresponding rendering logic.
Replace DEMO material with real content or clearly isolate it. Turn off a demo state only when the corresponding demonstration content is no longer shown.
Confirm consistency across titles, versions, sample lists, papers, and download packages. Describe research completion separately from public resource availability.
Test public links while signed out, and check mobile layout, previews, filters, details, and citation copying.
Leave content empty or entry points disabled when publication is unauthorized, materials are missing, or facts cannot be verified. See the full release checklist.
Maintenance summary: index.html controls content and layout; data/site-config.json controls site-wide resources and release information; data/samples.json controls the sample index. Maintain this page's HTML and Markdown together.