data_entry_ops · media · workflow
Labelbox helps Move.ai iterate 2x faster on computer vision algorithms with video annotation
Move.ai needed to train multiple AI algorithms for complex computer vision tasks — identifying people, classifying objects on a person such as jersey numbers, and recognizing human limbs in varied poses and distances, often occluded — across several distinct use cases, all requiring large amounts of labeled training data generated at speed.
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 · Training data need identified
Move.ai turned to Labelbox to spin up a labeling team, develop training data for their models, and iterate at top speed.
Tools used
LabelboxBoostPython SDKTensorflow · partnerPyTorch · partnerTensorRT · partnerNvidia RTX GPUsNvidia T4s
Outcome
Labelbox's data engine provided faster iterations on Move.ai's algorithms, helping them move 2x faster in the domain and accelerate their go-to-market strategy and subsequent product launches.
Results
Volume2x as fast
Grounding & classification
Source type: vendor customer story
26 fields verified against source quotes.
computer visiondata extractionhuman review describedmetric backednamed customerproduction runtime claimedsource backedtools describedvendor confirmedworkflow describedmediasoftwarecycle time reductiontime savedvendor customer storydata entry opshuman review queue