10.46243/jst.2021.v6.i3.pp169-177 registered
Deep Convolutional Generative Adversial Network on MNIST Dataset
Resolves to https://www.jst.org.in/index.php/pub/article/view/854
Held by Longman Publishers (India) · prefix 10.46243 live · DOI address https://doi.org/10.46243/jst.2021.v6.i3.pp169-177
Registered 29 Sep 2026 via crossref · record version 2 · last change 30 Sep 2026, 12:00 AM · record sha256 c9065cf8ea89c5f4…
Resolve ⬇ Record (JSON) ⬇ Kernel Metadata Declaration (XML) Compare with Crossref Cite (APA · BibTeX · RIS · CSL)
Also as schema.org JSON-LD · Turtle · RDF/XML · DataCite XML · Crossref UNIXREF · Crossref UNIXSD — the formats of the DOI content-negotiation service, from this record (which carries what).
What the DOI identifies
JournalArticle — an article in a journal · Digital · Visual · en
Deep Convolutional Generative Adversial Network on MNIST Dataset (PrincipalTitle · en)
Published 2021-06-02
Part of Journal of Science & Technology · ISSN 2456-5660 · volume 06 · issue 03 · pages 169–177
Agents
- S. Vijaya Lakshmi (author)
- Vallik Sai Ganesh Raju Ganaraju (author)
- Longman Publishers (publisher)
Identifiers DOI 10.46243/jst.2021.v6.i3.pp169-177
Abstract · en
In recent years, supervised learning with convolutional networks (CNNs) has seen huge adoption in computer vision applications. Comparatively, unsupervised learning with CNNs has received less attention. In this work we hope to help bridge the gap between the success of CNNs for supervised learning and unsupervised learning. We introduce a class of CNNs called deep convolutional generative adversarial networks (DCGANs), that have certain architectural constraints, and demonstrate that they are a strong candidate for unsupervised learning. Training on various image datasets, we show convincing evidence that our deep convolutional adversarial pair learns a hierarchy of representations from object parts to scenes in both the generator and discriminator. The generator uses tf.keras.layers.conv2Dtranspose (up sampling) layers to produce an image from a seed (random noise). Start with a dense layer that takes this seed as input, then up sample several times until you reach the desired image size of 28x28x1. The discriminator is a CNN-based image classifier. The model will be trained to output positive values for real images, and negative values for fake images. We define the Generator loss and the discriminator loss and we finally we get new images that look similar to our input(MNIST) images.
System metadata — ISO 26324:2025, Annex B · DOI Handbook 10.1
Each element by the standard's name (Annex B: reference elements, then administrative) and the Handbook's (in grey), read off the record above.
| Element | Value | In the record |
|---|---|---|
| DOI Name DOI name | 10.46243/jst.2021.v6.i3.pp169-177 | doi |
| Referent Type referentType | Creation | referent |
| Referent Sub-Type referentSubType | JournalArticle — an article in a journal | type |
| Referent Name(s) referentName(s) | Deep Convolutional Generative Adversial Network on MNIST Dataset (PrincipalTitle, en) | titles |
| Basic Metadata basicMetadata | author: S. Vijaya Lakshmi author: Vallik Sai Ganesh Raju Ganaraju publisher: Longman Publishers published: 2021-06-02 part of: Journal of Science & Technology · ISSN 2456-5660 · vol. 06 · no. 03 · pp. 169–177 language: en form: Digital · Visual · Language | agents, dates, container, language, structural_type, modes, characters |
| Referent Identifier(s) alternateIdentifier(s) | none besides the DOI | identifiers, relations (IsSameAs) |
| Registration Authority registrationAuthorityCode | Crossref — issued by Crossref (member 25296); held here as a copy | record.source_agency (our code, ra_doi_name, for names issued here once appointed) |
| Created Date issueDate | 2026-09-12 | record.registered (when the DOI name was first registered) |
| relatedIdentifiers | none needed — the descriptive metadata is in this record | container, relations (only where the descriptive metadata lives at another identifier) |
complete Every System Metadata element is here, with the basic metadata a journal article needs.
The System Metadata Declaration (JSON) · the Kernel Metadata Declaration (XML) · what each sub-type needs
Cited by — in this registry
No record held in this registry cites this DOI.
Counts only records held in this registry; a citation from a work held elsewhere is not seen here. The list as JSON
References with DOIs
The record lists 11 references; 3 carry a DOI.
| # | Reference | DOI | Held here |
|---|---|---|---|
| 5 | ref5 Dosovitskiy, Alexey, Springenberg, Jost Tobias, and Brox, Thomas. Learning to generate chairs with convolutional neural networks. arXiv preprint arXiv:1411.5928, 2014 | 10.1109/cvpr.2015.7298761 | not held here |
| 9 | ref9 Coates, Adam and Ng, Andrew Y. Learning feature representations with k-means. In Neural Networks: Tricks of the Trade, pp. 561–580. Springer, 2012 | 10.1007/978-3-642-35289-8_30 | not held here |
| 11 | ref11 Deng, Jia, Dong, Wei, Socher, Richard, Li, Li-Jia, Li, Kai, and Fei-Fei, Li. Imagenet: A large-scale hierarchical image database. In Computer Vision and Pattern Recognition, 2009.… | 10.1109/cvpr.2009.5206848 | not held here |
Relations
The record states no relation, and no record held here states one with this DOI.
Funding
The record names no funder.
Each funder's page lists every work held here that names it. ROR IDs open at ror.org; Funder Registry IDs resolve at doi.org.
History — the ledger
Every change to this DOI, in order, as it was recorded. Entries are only ever added, never changed or removed.
| # | When | What | By | Changes |
|---|---|---|---|---|
| 1 | 29 Sep 2026, 10:00 PM | register registered at Crossref; record read from api.crossref.org | Administrator (admin) | 83 fields set · sha256 bee4b5eab62a… |
| 2 | 30 Sep 2026, 12:00 AM | update record re-read from api.crossref.org | Administrator (admin) | container.titles.0.value: |
Machine-readable: the history as JSON, with the full record after each change.
Changes made at Crossref through DOI Doctor
Every deposit Smart Scholars' DOI Doctor made for this DOI, oldest first, with Crossref's own answer. Each fingerprint is the SHA-256 of its entry and of the fingerprint before it, so a change to any earlier entry changes every fingerprint after it.
| When | What changed | Crossref | Fingerprint |
|---|---|---|---|
| 13 Sep 2026, 01:44 | Address updated from https://www.jst.org.in/index.php/pub/article/view/854/781 to https://www.jst.org.in/index.php/pub/article/view/854 | ● accepted by Crossref | e6519aa1bf52… |
| 15 Sep 2026, 20:21 | Read back: the DOI resolves to the new address | ● confirmed | 4a07dfa3fc74… |
| 15 Oct 2026, 15:30 | Metadata completed abstract (202 words) · licence | ● accepted by Crossref | 6f2bc8a18ca1… |
| 15 Oct 2026, 16:20 | Read back: Crossref's record shows it | ● confirmed | d41a93fc3467… |
4 entries · the last fingerprint d41a93fc3467e5f9a45109d31a94217b14b44de93a292e68a8a14911ec28cf46 · the history as JSON, each entry with the fingerprint before it, to check them yourself.



