Low-Shot Learning from Imaginary Data
Citations Over TimeTop 1% of 2018 papers
Abstract
Humans can quickly learn new visual concepts, perhaps because they can easily visualize or imagine what novel objects look like from different views. Incorporating this ability to hallucinate novel instances of new concepts might help machine vision systems perform better low-shot learning, i.e., learning concepts from few examples. We present a novel approach to low-shot learning that uses this idea. Our approach builds on recent progress in meta-learning ("learning to learn") by combining a meta-learner with a "hallucinator" that produces additional training examples, and optimizing both models jointly. Our hallucinator can be incorporated into a variety of meta-learners and provides significant gains: up to a 6 point boost in classification accuracy when only a single training example is available, yielding state-of-the-art performance on the challenging ImageNet low-shot classification benchmark.
Related Papers
- → Improved Few-Shot Learning for Images Classification(2022)2 cited
- One Shot Device의 저장 신뢰도 분석에 관한 연구(2016)
- Does Use of Steel Shot Result in Greater Waterfowl Crippling? Using Harvest Data to Determine Effectiveness of Non-toxic Shot(2017)
- → One Shot Keto Full Reviews, Price, Benefits and How To Buy One Shot Keto?(2020)
- I've heard that there's a high-dose flu shot available. Is it better for older adults than the regular flu shot? Should I ask for the high-dose shot?(2011)