• MNIST: Handwritten Digits (unimodal)
• Fashion-MNIST: Zalando’s Article Images (unimodal)
• CIFAR-10/100: Tiny Natural Images Dataset (unimodal)
• CANDOR: Conversational Audio-Visual Dataset for Human Interaction (multimodal)
• Coswara: Breathing, Cough, and Speech Sounds for COVID-19 Diagnosis (unimodal - audio)
• ImageNet: Large-Scale Visual Recognition Dataset (unimodal - image)
• WIT: Wikipedia-based Image-Text Dataset (multimodal)
• AudioSet: Audio Events in YouTube Videos (unimodal - audio)
• VoxCeleb: Large-Scale Speaker Recognition Dataset (multimodal - audio & video)