Method

Image classification

Classifying images with CNNs or vision transformers.

20 cases from 14 companies apply Image classification. 4 were checked in an online A/B test; 6 report results in production.

Typical tasks: Catch fraud and abuse, Moderate content at scale, Detect anomalies, Optimize decisions and 7 more. Companies: DoorDash, Etsy, Wildberries, VK, Yelp, CFT and 8 more. Often combined with: U-Net, Human in the loop, Data augmentation, Threshold selection, Gradient boosting and 49 more.

With Eyes Wide Shut: Detecting Sleeping People and Other Remote Verification Attacks Without Consentfintech2026-05
A Digital Core Documentermanufacturing2026-05
Digital Twins of the Heart and Speeding Them Uphealthtech2026-05
SafeChat: DoorDash's AI-powered safety featurefood-delivery2025-10
High-Speed Multimodal Moderation: LLMs and VLMs in Real Productionmarketplace2025-05
Machine Learning in Content Moderation at Etsymarketplace2024-08
From Small to Great: Finding and Growing Promising New Authors on VKsocial2024-05
Moderating Inappropriate Video Content at Yelpmarketplace2024-03
Seismic Data Through a ResNet18 Lensmanufacturing2023-05
Diagnosing Oil and Gas Pipelines Using Machine Learning—manufacturing2023-05
A Review of Artifacts in Histopathology Images—healthtech2023-05
Solving the Face Spoofing Problem in Photographsfintech2023-05
From Image Classification to Multitask Modeling: Building Etsy's Search by Image Featuremarketplace2023-05
Automated profile picture moderation at BlaBlaCar using Deep Learningmobility2023-03
How DoorDash Upgraded a Heuristic with ML to Save Thousands of Canceled Ordersfood-delivery2023-01
When a picture is worth more than wordstravel2022-12
Generation of three-dimensional synthetic data for training neural networksgeneral2022-10
What is good for a doctor is death for an engineer: extra requirements for medical AI services—healthtech2022-10
What is good, what is bad. Quality of X-ray images—healthtech2022-10
Poster generation for films and cartoons in an online cinemamedia-streaming2022-06