Manufacturing & Agriculture

Cleanlab ensures reliable data/models in industrial quality control applications (and more generally across enterprise computer vision, natural language processing, and statistical data mining). Use Cleanlab Studio to seamlessly handle mislabeled data, outliers/anomalies, drift, ambiguous instances, and other real-world data issues.
Hero Picture

Case StudyAutomatically Correcting Image Labels

error improvement for ResNet computer vision model (without any change in modeling code)
Quote from Travis Tang, Data Scientist at Gojek:
I used an open-sourced library, cleanlab, to remove low-quality labels on an image dataset. The [ResNet] model trained on the dataset without low-quality data gained 4 percentage points of accuracy compared to the baseline model (trained on all data).
Gojek is an Indonesian on-demand multi-service platform and digital payment technology group.
Company Logo

Case StudyCleanlab Finds Thousands of Errors in ImageNet

ImageNet is the most famous computer vision (image recognition) dataset with millions of images. Cleanlab Studio automatically found tens of thousands of data errors like label issues, outliers, ambiguous examples, and (near) duplicates. The graphic below shows a few of them. Read more.

Your Picture

Browse other labeling errors detected by Cleanlab in famous ML benchmark datasets at


Automatically detect potential issues in image datasets like images that are:
  • under/over-exposed
  • blurry
  • near duplicates
  • low-information
Videos on using Cleanlab Studio to find and fix incorrect labels for:
Automatically detect outliers (anomalies) which may have an outsized impact on data-driven conclusions and should be handled with care. Learn more.
Model images together with tabular (numeric, categorical) and text information.
Know which subset of the data is high-quality with confidence, and evaluate the quality of different data sources.
Effectively analyze data labeled by multiple annotators, and estimate which examples require additional review and which annotators are best/worst overall. Learn more.
Use our ActiveLab system (active learning with relabeling) to efficiently collect new labels for training accurate models.
Read more about why 2022 was the most exciting year in computer vision history and how Cleanlab fits into it.
Read more about why the foundations of AI are riddled with errors.
Automatically identify and resolve data issues, and deploy robust ML models with a few clicks. Cleanlab Studio facilitates data-centric AI workflows in:
  • Agricultural applications: disease inspection, yield estimation, animal monitoring, as well as tasks involving grading and sorting.
  • Industrial quality control applications: ingredient inspection, process quality monitoring, assembly inspection, and defect detection.