Stanford Cars
A frozen DINOv2 ViT-S/14 backbone with a dictionary of 1,024 concepts, up to 16 per image, and 16 quantized attributes of 64 values each. Across its 8,144 training images, a typical image has 16.
Cadillac Escalade EXT Crew Cab 2007 · Bentley Continental GT Coupe 2007 · Chrysler Crossfire Convertible 2008
Chevrolet Silverado 1500 Extended Cab 2012 · Chevrolet Silverado 2500HD Regular Cab 2012 · Chevrolet Silverado 1500 Regular Cab 2012
GMC Savana Van 2012 · Dodge Sprinter Cargo Van 2009 · Chevrolet Express Van 2007
Volvo 240 Sedan 1993 · Volkswagen Golf Hatchback 1991 · Audi V8 Sedan 1994
Chevrolet Silverado 1500 Classic Extended Cab 2007 · Chevrolet Silverado 1500 Hybrid Crew Cab 2012 · Dodge Dakota Club Cab 2007
Ferrari 458 Italia Coupe 2012 · Bugatti Veyron 16.4 Coupe 2009 · Lamborghini Aventador Coupe 2012
Ford F-450 Super Duty Crew Cab 2012 · Cadillac Escalade EXT Crew Cab 2007 · Dodge Durango SUV 2007
Volvo C30 Hatchback 2012 · Audi 100 Wagon 1994 · Chevrolet Avalanche Crew Cab 2012