Technology

Transfer learning

Repurpose pre-trained models (like ResNet or BERT) to solve new tasks with minimal data and compute.

Transfer learning bypasses the need for massive datasets by leveraging weights from models trained on massive corpuses: ImageNet for vision or Wikipedia for NLP. Instead of training 100 million parameters from scratch, you freeze the base layers and fine-tune the final classification head. This approach cuts training time by up to 90% and enables high-performance AI on niche datasets (such as identifying rare medical anomalies) where only a few hundred samples exist. Industry standards like BERT and EfficientNet rely on this logic to deliver state-of-the-art results without the massive energy costs of full-scale training.

https://www.tensorflow.org/tutorials/images/transfer_learning

What builders pair with Transfer learning

Projects using both technologies. Select a pairing to see a project.

Built with Transfer learning

CUDA 1 shared project Deep learning 1 shared project diffprivlib 1 shared project Dropout 1 shared project FER2013 1 shared project FFmpeg 1 shared project

12 more pairings

Generative AI 1 shared project Google Differential Privacy Library 1 shared project GStreamer 1 shared project H 1 shared project Kaggle 1 shared project Keras 1 shared project llama 1 shared project Mixup 1 shared project NVIDIA DeepStream 1 shared project OpenCV 1 shared project OpenDP 1 shared project Pandas 1 shared project

Pairing: CUDA

Seguridad Infantil: El Uso de Cámaras de Video e IA para el Cuidado de Niños

Bogotá · November 28, 2024

view project | view CUDA

Recent Talks & Demos

Showing 1-4 of 4

Members-Only

Sign in to see who built these projects

Sign in View FAQ

Modèle CNN comme Deep Learning pour la reconnaissance des expressions…

ResNet-18 ResNet-50

IA en Seguridad y Cuidado Infantil

RNN Transfer learning

Convnets for Semiconductor Identification