Context-Aware Meta-Learning
This addresses the challenge of enabling visual models to learn new objects during inference, similar to large language models, for applications in computer vision and meta-learning.
The paper tackles the problem of visual models learning new concepts during inference without fine-tuning, proposing a meta-learning algorithm that matches or exceeds state-of-the-art performance on 8 out of 11 benchmarks without meta-training or fine-tuning.
Large Language Models like ChatGPT demonstrate a remarkable capacity to learn new concepts during inference without any fine-tuning. However, visual models trained to detect new objects during inference have been unable to replicate this ability, and instead either perform poorly or require meta-training and/or fine-tuning on similar objects. In this work, we propose a meta-learning algorithm that emulates Large Language Models by learning new visual concepts during inference without fine-tuning. Our approach leverages a frozen pre-trained feature extractor, and analogous to in-context learning, recasts visual meta-learning as sequence modeling over datapoints with known labels and a test datapoint with an unknown label. On 8 out of 11 meta-learning benchmarks, our approach -- without meta-training or fine-tuning -- exceeds or matches the state-of-the-art algorithm, P>M>F, which is meta-trained on these benchmarks. Our code is available at https://github.com/cfifty/CAML.