back_office_ops · workflow

How Dropbox built AI-powered image content search using EfficientNet and word vectors

Photos in Dropbox are difficult to find by filename because camera-generated filenames contain no content information, forcing users to browse thumbnails manually to locate specific images.

How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · Image classified on upload
An EfficientNet network trained on the OpenImages dataset classifies each image and produces scores for about 8,500 categories.
Tools used
EfficientNetOpenImagesTensorFlowConceptNet Numberbatchword2vecNautilusOCR
Outcome

Image content search is now enabled for all Professional and Business Dropbox users, combining image classification with OCR and full-text search; index storage was reduced to 500 bytes per image from 80 kilobytes, and query latency is on par with text search.

Results
Time saved10 posting lists — roughly the same amount of work we do for text queries
Volume500 bytes instead of 80 kilobytes
Source

https://dropbox.tech/machine-learning/how-image-search-works-at-dropbox

How we source this →

Grounding & classification
Source type: technical build writeup
23 fields verified against source quotes.
computer visionenterprise searchocrbuilder submittedmetric backedproduction runtime claimedtools describedworkflow describedsoftwarecost reductioncycle time reductiontechnical build writeupback office opsextract classify route