Selective Zero-Shot Classification with Augmented Attributes
This addresses the issue of unreliable predictions in zero-shot classification for AI systems, though it is incremental as it builds on existing attribute-based methods.
The paper tackles the problem of selective zero-shot classification, where classifiers avoid dubious predictions, by proposing a method that uses both human-defined and automatically discovered residual attributes to improve performance. Experiments show it outperforms other methods in risk-coverage trade-off metrics.
In this paper, we introduce a selective zero-shot classification problem: how can the classifier avoid making dubious predictions? Existing attribute-based zero-shot classification methods are shown to work poorly in the selective classification scenario. We argue the under-complete human defined attribute vocabulary accounts for the poor performance. We propose a selective zero-shot classifier based on both the human defined and the automatically discovered residual attributes. The proposed classifier is constructed by firstly learning the defined and the residual attributes jointly. Then the predictions are conducted within the subspace of the defined attributes. Finally, the prediction confidence is measured by both the defined and the residual attributes. Experiments conducted on several benchmarks demonstrate that our classifier produces a superior performance to other methods under the risk-coverage trade-off metric.