/projects/video-classification-metadata-generation-sih
Video Classification & Metadata Generation (SIH)build dossier
PythonTensorFlowOpenCVPandasDeep Learning
build.metadata
RoleDeveloper — Smart India Hackathon
DurationMar 2022 – Jun 2022
StatusCompleted
TelemetryActive
overview
An ISRO problem statement built for the Smart India Hackathon: a Python deep-learning pipeline using CNNs and RNNs to interpret video frames, classify the video, and generate metadata as CSV. Built with OpenCV, Pandas, and TensorFlow.
implementation.notes
1Design the interface around repeatable workflows and inspectable states.
2Keep the stack typed end-to-end so APIs, data models, and UI contracts stay aligned.
3Use deployment-friendly boundaries between frontend, backend, persistence, and integrations.
readonlyconnected
design.principles
Product surface
A focused interface layer for the user-facing workflow.
Stack boundary
Technologies grouped around UI, data, and deployment concerns.
Maintainability
Designed to be inspectable, typed, and practical to extend.
Want to discuss a similar build?
Use this project as a reference point and send the workflow, timeline, and constraints you care about most.