SciViD: The Publisher of Video Science

OUR VISIONScholarly publishing remains essential for disseminating knowledge, but it faces persistent problems of cost, access, transparency and format that limit who can participate and how research is communicated. SciViD is being developed as a complementary publishing model that combines video presentation, AI-assisted production and an inclusive editorial framework, with the aim of addressing some of these problems.CHALLENGES IN SCHOLARLY PUBLISHINGCost and access. Subscription journals restrict access through paywalls, and many open-access journals charge article processing charges (APCs). Both can disadvantage early-career researchers, students and institutions with limited funding.Opaque peer review. Peer review varies in quality, can be affected by bias, and is rarely visible to readers.Unequal participation. Scholarly publishing is concentrated in a small number of regions, and researchers who are not native English speakers face additional barriers to publication.Limits of static formats. Text and static figures convey dynamic processes, experimental technique and simulations poorly, which can hinder reproduction and teaching.HOW SCIVID RESPONDSOpen access without high charges. SciViD will charge no article processing charges during its initial phase and is developing a funding model intended to keep video-based open-access publishing affordable.Traceable review. Peer review will be managed in an auditable system that records each manuscript’s review history, and every published video paper will be labeled with the review track it passed through.Multilingual access. SciViD is developing multilingual subtitles and AI-assisted translation to widen access for researchers and readers who are not native English speakers.Visual presentation of methods. Video papers present methods, simulations and devices visually rather than in text alone. This format is intended to support reproduction and to make published work usable for teaching.AI-assisted production. AI tools for narration, animation, editing and translation are intended to reduce the cost and effort of producing a video paper. All AI use must be disclosed, and AI-generated visuals must be labeled in accordance with SciViD’s Integrity Guide.Pathways for early-career researchers. The Emerging Scholars track offers pre-university students and early-career researchers a supervised route to producing methodologically sound, publishable research.

LIMITATIONS OF STATIC PUBLICATION FORMATSThe text-based research paper remains the principal record of scientific work, and its precision and archival stability are not in question. Text and static figures are, however, poorly suited to conveying certain kinds of information: practical technique, dynamic processes, and the course of an experiment or simulation over time. SciViD’s video papers are intended to address these limitations.1. Under-reported practical techniqueDetails that determine whether an experiment can be repeated, such as instrument settings, timing, sample handling and manual technique, are difficult to specify fully in prose. In the Reproducibility Project: Cancer Biology, none of the experiments selected for replication was described in enough detail to design a replication protocol without contacting the original authors. Incomplete methods reporting is one of several causes of replication failure, alongside publication bias, incomplete statistical reporting, and limited sharing of data and materials; visual documentation of technique addresses only the first of these.2. Dynamic processes reduced to snapshotsProcesses involving motion or transformation, such as fluid flow, self-assembly or the behavior of living cells, must be reconstructed by the reader from a sequence of still images. Research on instructional animation shows a moderate overall learning advantage of animation over static pictures, with larger effects for realistic, video-based material and for learning procedural skills. The benefit is not automatic: dynamic formats help mainly when they show change that static images cannot, and when pacing and signaling are well designed. These findings concern learning outcomes; whether they translate into higher replication success has not, to our knowledge, been tested directly.3. Barriers for readers outside the specialtyResearch papers are written for specialists, and their terminology and compressed style limit access for students, researchers from adjacent fields and non-native English speakers. Video papers that pair narration with visual explanation can lower some of these barriers, because people generally learn better from words and pictures together than from words alone.4. Video as a supplement rather than the recordMany journals already accept supplementary videos, and dedicated video journals exist. In conventional papers, however, video is usually supplementary material rather than the primary object of review and citation. SciViD treats the video paper itself as the reviewed and citable record.THE PATH FORWARDText-based papers will remain the foundation of the scholarly record. Formats that show what text describes, including video papers, interactive simulations, and shared data and code, can complement them by making methods easier to see, evaluate and repeat.

Text-based scientific papers continue to serve as the cornerstone for archival precision. However, in an evolving digital landscape, the limitations of static formats in conveying complex experimental processes and ensuring reproducibility have become increasingly apparent. SciViD is being developed as a complementary publishing model intended to preserve academic rigor while supporting new modes of scientific expression, dissemination, and participation.1. Scientific Rigor with Multimedia Integration SciViD is designed to preserve the core principles of scholarly publishing while making full use of the expressive capacity of video:Editorial Standards: A peer-review and editorial framework, currently in development, is intended to support scientific validity and methodological clarity.Persistent Identification: Planned DOI registration through Crossref will provide stable citation and support future indexing in academic databases.Structured Scholarship: Submissions are structured to align with familiar scholarly conventions (Abstract, Introduction, Methods, Results, etc.).Visual Evidence: SciViD enables visual demonstrations of dynamic processes, narrated walkthroughs of experimental setups, and high-fidelity simulations.This hybrid format is particularly well suited to fields where movement and transformation are central, such as microfluidics, nanotechnology, quantum physics, and experimental biology.2. Accessibility and Multilingual Inclusivity SciViD is designed to reduce common barriers to scientific participation:Global Reach: Planned support for multilingual subtitles and AI-assisted translation is intended to improve accessibility for non-native English-speaking readers.Inclusive Design: Multimodal presentation (visual + auditory) is intended to support a range of learning preferences. Realizing genuine accessibility benefits for users with visual or auditory impairments specifically will depend on providing captions, transcripts, and audio description alongside video content.Digital Literacy: The platform is being built for mobile and web access, with the aim of reaching researchers and students across different regions. Because video is more bandwidth-intensive than text, meeting this goal in lower-bandwidth settings may require lower-resolution or downloadable viewing options.3. Pedagogical Optimization and Student Empowerment Unlike traditional journals, SciViD is designed with educational utility in mind, giving students a pathway from passive consumers of science to active contributors:Project-Based Learning: The video format is well suited to flipped classrooms and project-based digital curricula.Portfolio Development: Students can build verifiable skills in scientific communication and multimedia production through a portfolio of peer-guided work.Mentorship and Feedback: Educators and mentors may provide informal feedback and guidance as part of the learning process, separate from SciViD's independent, formal peer-review process, which is conducted without their involvement.4. Methodological Transparency and Instrument Literacy SciViD is designed to help address a recognized gap in traditional literature by enabling visual documentation of experimental protocols:Direct Observation: Viewers can see instrument calibration, sample handling, and troubleshooting steps directly, as they were performed, rather than inferring them from written description.Reproducibility: Providing visual documentation of tacit, procedural knowledge is intended to complement written protocols for highly specialized or safety-sensitive procedures in chemistry and engineering, where written description alone is often insufficient for reliable replication.5. AI-Native Architecture and Future Integration SciViD submissions are being designed for compatibility with emerging AI-based knowledge systems:Semantic Search: Content is being designed for AI transcription and segmentation, with the goal of enabling indexing within AI tutoring systems and knowledge graphs as these tools mature.Structured Metadata: Planned deep metadata tagging is intended to support concept-based retrieval, beyond simple keyword search.Simulation Extensions: The platform is designed to eventually support AR/VR overlays and physics-based simulations.6. Ethical and Sustainable Publishing Economics SciViD is developing a publishing model intended to address the high costs associated with traditional open-access publishing:Low Barriers: SciViD charges no article processing charges (APCs) during its initial phase, removing a major cost barrier for early-career researchers, students, and underfunded institutions.Value Recognition: We are developing frameworks intended to fairly recognize contributors — authors, reviewers, and institutions — for their contributions.Institutional Partnership: Schools and universities can integrate SciViD into their teaching and research activities.7. Hybrid Scholarly and Public Visibility SciViD is designed to support both scholarly and public-facing visibility for published work:Dual Presence: Articles are designed to reside within both scholarly archives and professional video platforms, to support broader reach.Discoverability: Once DOI registration and structured metadata are in place, published work is intended to be findable through search engines and academic discovery tools such as Google Scholar.Cross-Channel Impact: A single SciViD publication is intended to serve as a formal research artifact, a classroom resource, and supporting material for funding applications or conference presentations.SciViD's aim is to combine established scholarly standards with visual communication, supporting closer engagement with experimental science and a more accessible model of authorship for early-career researchers.

SciViD offers two distinct submission pathways, tailored to the author's academic goals and experience level. Both pathways prioritize visual clarity and methodological transparency, but differ in review depth and institutional positioning.1. The Emerging Scholars Collection (Editorial Review + Planned DOI)Purpose: Designed for high-achieving high school and undergraduate students, science clubs, and early-career researchers building a research portfolio.Focus: Mentorship, structured scientific thinking, and the creation of a verifiable, citable research artifact.Review Process: Undergoes Supervised Editorial Review. Each submission is reviewed by a SciViD PhD mentor not involved in supervising that student's project, assessing clarity, ethical AI use, and methodological soundness. It is an instructional review process designed to help bring the student's work toward professional standards.DOI & Indexing: Planned permanent DOI (via Crossref) and a dedicated metadata page, intended to make the work a stable, citable record once implemented.Recognition: Authors are credited as Junior Research Scholars.Best for: Students seeking a structured, mentored research experience that produces a citable, verifiable output.2. The Flagship Journal (Formal Peer Review + Planned DOI)Purpose: Designed for senior researchers, PhD candidates, and professors seeking formal academic recognition and contribution to the global body of scientific knowledge.Focus: Original research findings, comprehensive literature reviews, and high-fidelity experimental simulations.Review Process: Undergoes Formal Peer Review. Submissions are evaluated through a structured editorial workflow and, as our network expands, independent peer assessment, to assess scientific validity.DOI & Indexing: Planned permanent DOI and standardized metadata, intended to support discovery in academic databases and library systems once implemented.Recognition: Published as a Peer-Reviewed Video Article.Best for: Researchers and educators aiming to publish video-based research with transparent, method-centered documentation.SUMMARY OF KEY DIFFERENCESFeatureEmerging Scholars CollectionFlagship Journal PublicationPrimary AudienceHigh School / UndergraduateProfessional Researchers / PhDsReview TypeSupervised Editorial ReviewFormal Peer ReviewCore GoalPortfolio & Skill DevelopmentAcademic Contribution & CitationDOI RegistrationPlanned (via Crossref)Planned (via Crossref)IndexingDiscovery-optimized MetadataAcademic Database OptimizationAuthor StatusJunior Research ScholarAuthorPlatformSciViD Digital LibrarySciViD Flagship JournalAN INTEGRATED ECOSYSTEMBy offering these two pathways, SciViD aims to connect science education with professional publishing: supporting early-career researchers in developing structured, citable research experience, while giving established researchers a method-centered, video-based medium for transparent scientific communication.

Yes. SciViD welcomes original computational modeling, numerical analysis, and simulation research. Theoretical and in-silico investigation is central to modern scientific progress, and SciViD is designed to showcase the "how" and "why" of complex digital experiments through video.ACCEPTABLE TYPES OF MODELING & SIMULATION PAPERS1. Original Research SimulationsPapers based on original numerical investigations where the author's code or modeling environment (e.g., MATLAB, COMSOL, Python, ANSYS, LAMMPS) produced novel data.Requirement: Authors must explain the methodology, input parameters, and validation process. Visualizations should be derived directly from simulation outputs.2. AI-Driven & Predictive ModelingPapers utilizing machine learning or neural networks to predict phenomena (e.g., protein folding, material behavior, or climate patterns).Requirement: Authors must present the datasets, training methodology, and validation metrics. As with the numerical-simulation track, every visual must be labeled to distinguish the model's actual predicted output (e.g., a rendering generated directly from the trained model's predictions) from any separately AI-generated illustrative or explanatory content added around it.3. Methodological & Conceptual ReviewsEducational or review papers that explain existing models or complex theoretical frameworks.Requirement: While the author may use conceptual animations (e.g., Manim, Blender) or metaphoric visuals to aid understanding, all original sources must be cited, and visuals must be labeled as "conceptual representations."ACADEMIC INTEGRITY & PROHIBITED PRACTICESTo maintain rigorous standards of scientific communication, SciViD prohibits:Pseudo-Simulations: Presenting AI-generated animations or "faked" visuals as if they were generated by a scientific solver when they were not.Misleading Visuals: Animations that suggest experimental behavior without being grounded in underlying data or cited theory.Unattributed IP: Utilizing simulation code, datasets, or visual outputs created by others without proper citation or legal authorization.SUBMISSION REQUIREMENTSTo ensure a Modeling or Simulation paper meets SciViD standards, the following must be included:Methodological Transparency: Clear description of the software, coding environment, governing equations, boundary conditions, and computational setup.Visual Integrity: Visuals must accurately reflect simulation data. Every frame must be labeled to distinguish between Empirical Simulation Output and Conceptual Animation.Open Science (Recommended): We strongly encourage the sharing of source code, configuration files, or links to public repositories (e.g., GitHub) to support reproducibility.Mandatory AI Disclosure: If any part of the video — including narration, background visuals, or specific animations — was generated or enhanced by AI, this must be explicitly stated in the submission metadata and the video credits.Educational Narrative: The video should not only present data but also provide a narrative "walkthrough" explaining the significance of the model, the logic of the parameters, and the broader context of the results.EXAMPLE TITLES IN THE SIMULATION TRACK"Visualizing Magnetic Field Lines in Tokamak Reactors Using Python and OpenGL""Simulating Urban Heat Islands: A GIS and Agent-Based Modeling Approach""Modeling Quantum Tunneling in Nanowires with Time-Dependent Schrödinger Equation""Predicting Drug-Target Interactions with a Deep Learning Framework (TensorFlow-based)""Educational Review: Molecular Dynamics of Lipid Bilayers Visualized with LAMMPS and VMD"SciViD is designed to turn computational work into dynamic, accessible science communication. Whether you are modeling fluid flows or training neural networks, the aim is to make your work not just recorded, but genuinely observable — and easier for others to attempt to reproduce.

At SciViD, we aim to combine rigorous scholarship with structured video presentation. Because the platform accepts submissions in both Experimental Video Papers (empirical data) and Review/Conceptual Video Papers (literature synthesis), a "zero-distort" policy applies across both. This guide sets out requirements intended to ensure that AI-generated media enhances understanding without compromising academic integrity or misleading the scientific community.ONE PLATFORM, THREE VISUAL STANDARDS1. Experimental Video PapersDefinition: Original research documenting empirical experiments performed by the author(s).Requirement: Must center on authentic video footage, micrographs, sensor data, or graphs recorded during the actual research process.Role of AI: Limited to cosmetic enhancements (e.g., UI layout, titling, background denoising, or transition effects). 🚫 Strictly Prohibited: Using AI to "fill in" missing experimental steps, simulating results that weren't captured, or using AI-generated stock footage to represent original data.2. Review & Conceptual Video PapersDefinition: Educational content or syntheses based on existing literature.Requirement: Authors are encouraged to use AI-generated diagrams, whiteboard animations, and visual metaphors to explain complex concepts.Role of AI: High creative freedom to aid pedagogy.Requirement for Transparency: Every AI-generated asset must be clearly labeled as "Conceptual Animation" or "Illustrative Metaphor." It must never be presented as empirical evidence.3. Computational & Predictive Research PapersDefinition: Original research where a trained model (e.g., machine learning, neural network) produces the paper's core predictive or generative findings, as covered under SciViD's Modeling & Simulation submission track.Requirement: Visuals representing the model's actual predicted output must be clearly labeled as "Model-Predicted Output," distinct from any separately AI-generated illustrative material used to explain the method.Role of AI: The model's output is the paper's data. Any additional AI-generated content used purely for explanation or presentation follows the same "Conceptual Animation" / "Illustrative Metaphor" labeling used in Review & Conceptual Video Papers. 🚫 Strictly Prohibited: Presenting an illustrative AI-generated rendering as though it were the model's actual predicted output, or vice versa.THE 6 CORE PRINCIPLES OF AI INTEGRITY1. Explicitly Distinguish Fact from Illustration Never leave the viewer guessing. If a visual is an AI-generated artist's interpretation rather than raw data or model output, it must be declared via on-screen labels, audio cues, or the video description. SciViD's standard label set: "Conceptual Animation," "Illustrative Metaphor," "Hypothetical Scenario," and "Model-Predicted Output."2. Zero Tolerance for Data Fabrication AI should never be used to "hallucinate" results. Even in educational reviews, it is misconduct to:Create visuals suggesting false data trends.Modify published figures from other authors to support a different conclusion.Show imagined experimental setups without a clear "Hypothetical Scenario" label.3. The Full "Tool-Stack" Disclosure Every SciViD submission must include an AI Use Statement detailing:The Tools: (e.g., Midjourney, Sora, ElevenLabs, Pika).The Scope: Which parts are AI (e.g., "Narrative voiceover," "Conceptual protein-folding animation").The Intent: Whether the AI was used to enhance the presentation of authentic content, to create illustrative or conceptual material, or to render a model's predicted output.4. Respect for Intellectual Property (IP) AI is not a license to bypass copyright. Do not use AI to replicate proprietary instrument designs, copyrighted figures from other journals, or trademarks without legal authorization. SciViD supports ethical, original creation.5. Precision Over Sensationalism Science is spectacular enough on its own — don't let AI turn it into sci-fi. Avoid "clickbait" visuals: cells shouldn't look sentient, and black holes shouldn't appear inside test tubes. Focus on clarity and accuracy, not cinematic exaggeration.6. Mandatory Alignment with Cited Sources Every AI-generated visual in a Review Video must strictly correlate with the sources it cites. If a source describes a reaction as "slow and subtle," the AI animation shouldn't show it as an "instant explosion."SCIVID'S GOLDEN RULEAI visuals must illuminate, not manipulate. They are tools for enhancing human understanding, never for simulating reality without full disclosure.SUBMISSION CHECKLIST: AI VISUAL INTEGRITY Is every AI-generated asset clearly labeled as "Conceptual Animation," "Illustrative Metaphor," "Hypothetical Scenario," or "Model-Predicted Output," as applicable? Have I ensured no AI was used to fabricate or approximate experimental outcomes or model results? Is the AI Use Statement included in the credits and metadata? (Review/Conceptual papers) Do all AI visuals align strictly with the sources cited? Have I verified that no AI content violates third-party copyrights or trademarks?Questions regarding visual ethics? Reach out to our editorial team at editor@scivid.ai.

SciViD is a video-based academic publisher that helps students and researchers turn their scientific work into video papers, reviewed through either Supervised Editorial Review or Formal Peer Review depending on the submission track. As generative AI becomes a staple in creative workflows — from 3D animation to synthetic narration — maintaining a foundation of responsibility, transparency, and ethics is essential.This guide outlines SciViD's standards for using generative AI tools — such as text-to-image (DALL·E, Midjourney), whiteboard animation (InstaDoodle), and text-to-video (Pika, Runway, Sora) — and works alongside SciViD's Integrity Guide, which sets out how AI-generated visuals must be labeled and disclosed.SciViD Ethical Guidelines on Generative AI1. Mandatory Declaration and Disclosure All submissions must explicitly identify which components were generated or assisted by AI.Scope of Disclosure: This includes AI-generated imagery, animations, synthetic voiceovers, automated script polishing, and translations.The AI Use Statement: Every video must include the AI Use Statement specified in SciViD's Integrity Guide (Tools / Scope / Intent) in the credits or metadata. Example: "Visuals created via Midjourney (illustrative); narration assisted by ElevenLabs Voice AI; script authored by the researcher with structural polishing via Claude 3.5."Visual Labeling: Where an AI-generated visual represents anything other than authentic recorded data, it must also carry the corresponding on-screen label defined in the Integrity Guide (Conceptual Animation, Illustrative Metaphor, Hypothetical Scenario, or Model-Predicted Output, as applicable).2. Authorship and Intellectual ContributionAI is a tool — not an author. While AI can assist with production, the core intellectual work of the paper — the hypothesis, experimental method, data analysis, and original conclusions — must be that of the named human authors.Student Autonomy: Students must remain the primary architects of their reasoning and narratives.Attribution: All mentors, teachers, or collaborators who assisted in the AI workflow must be transparently acknowledged.3. Privacy, Consent, and Copyright Integrity SciViD maintains a strict policy regarding the likeness and property of others. We do not accept:Deepfakes: AI-generated replicas of real people without explicit, documented consent.Copyright Infringement: Content generated using protected data without the legal right to reproduce it.Misrepresentation: AI-generated voices or images used to defame or misidentify individuals or institutions.4. Safety and Bias MitigationAll AI-generated content undergoes ethical screening. We prohibit:Inappropriate Content: Visuals depicting violence, hate speech, or illegal activities.Algorithmic Bias: Stereotypical or biased depictions of race, gender, culture, or religion.Misinformation: AI-generated "hallucinations" presented as scientific facts or verified data.5. Support, Don't Replace, Scientific ThinkingAI should function as your lab assistant, not your lead scientist. SciViD encourages generative tools to:Visualize the Invisible: Animate molecular interactions, cellular dynamics, or quantum states.Enhance Accessibility: Simplify complex concepts for public engagement.Improve Literacy: Assist non-native speakers in refining their scientific communication.Limit: Do not let AI dictate your experimental results or research direction.Alignment with National FrameworksFor school-aged authors, this guide is informed by the principles of the Australian Framework for Generative AI in Schools (approved 2023; reviewed 2024–25), prioritizing:Equity of Access: Aiming to use AI tools to narrow, rather than widen, the digital divide.Data Privacy: Protecting student identity and intellectual property from exploitative data harvesting.Critical AI Literacy: Teaching students to question the outputs, biases, and limitations of the tools they use.The SciViD Integrity AgreementStandardAuthor ResponsibilityHonest AttributionDisclose every tool used in the production process.No PlagiarismEnsure AI is not used to "paraphrase" existing works without citation.VerificationHuman authors are solely responsible for the accuracy of AI-generated claims.ConsequencesViolations may result in manuscript rejection, retraction, or institutional notification.Why SciViD Champions Ethical AIAI offers a genuine opportunity to make scientific publishing more visual and more widely understandable. That potential is only realized when paired with rigorous ethical standards.Questions regarding your AI workflow? Reach out to our editorial board at editor@scivid.ai.

Dear Colleagues, Educators, and Students,As a researcher and innovator working at the intersection of nanotechnology, microfluidics, and interdisciplinary science, I have spent the past two decades navigating the structural limitations of traditional academic publishing — both as an author and a mentor. My journey has taken me through world-renowned institutions, including a tenured professorship at Fudan University, postdoctoral research at the Max Planck Institute, and a visiting scholarship at Harvard University.These experiences led me to a pivotal realization: while science is becoming increasingly visual, dynamic, and collaborative, the way we share it remains static, text-heavy, and often inaccessible.In my own research — particularly with nanoscale engines, fluidic systems, and self-assembling materials — I frequently struggled to convey core discoveries through words and 2D images alone. I witnessed firsthand how students and colleagues could misinterpret protocols that were crystal clear in the laboratory but poorly communicated in print.More importantly, I have mentored many brilliant young scientists whose ideas deserved global attention but were sidelined by institutional gatekeeping, high journal fees, or linguistic bias. They could demonstrate the science, but they could not easily publish it. SciViD was created to change this.Our vision is to build a new kind of platform: one that is designed to retain the essential rigors of peer review and DOI indexing, while harnessing the expressive power of video to make science more understandable and reproducible. Whether it is an Emerging Scholar presenting their first experiment, or a global expert visualizing quantum concepts through animation, SciViD invites contributions that show science rather than just describe it. We are building SciViD around three core principles:Integrity: Applying strict editorial standards and transparent ethics to every publication.Inclusion: Supporting youth authorship and non-native English speakers through AI-enhanced accessibility features now in development.Impact: Working toward a sustainable ecosystem where contributors and institutions are recognized for their educational and scientific value.Our goal is not to replace traditional journals, but to complement them — bridging the gap between formal research and global learning. SciViD is intended as a tool for education: enabling teachers to use video-based materials once published works are available, helping students build an academic portfolio early in their careers, and giving institutions the option to adopt multimedia content for enrichment.In the long term, I envision SciViD as a living archive of scientific practice — where motion, method, and meaning converge, and where science is no longer locked behind PDFs and paywalls, but shared through voice, visuals, and verifiable experimentation.I warmly invite you to join us in building a new way to share and publish science.Professor Alexander A. SolovevFounder, SciViD – The Publisher of Video Science