Unsupervised Domain Expansion for Visual Categorization

Expanding visual categorization into a novel domain without the need of extra annotation has been a long-term interest for multimedia intelligence. Previously, this challenge has been approached by unsupervised domain adaptation (UDA). Given labeled data from a source domain and unlabeled data from a target domain, UDA seeks for a deep representation that is both discriminative and domain-invariant. While UDA focuses on the target domain, we argue that the performance on both source and target domains matters, as in practice which domain a test example comes from is unknown. In this article, we extend UDA by proposing a new task called unsupervised domain expansion (UDE), which aims to adapt a deep model for the target domain with its unlabeled data, meanwhile maintaining the model’s performance on the source domain. We propose Knowledge Distillation Domain Expansion (KDDE) as a general method for the UDE task. Its domain-adaptation module can be instantiated with any existing model. We develop a knowledge distillation-based learning mechanism, enabling KDDE to optimize a single objective wherein the source and target domains are equally treated. Extensive experiments on two major benchmarks, i.e., Office-Home and DomainNet, show that KDDE compares favorably against four competitive baselines, i.e., DDC, DANN, DAAN, and CDAN, for both UDA and UDE tasks. Our study also reveals that the current UDA models improve their performance on the target domain at the cost of noticeable performance loss on the source domain.

Boje u rečniku „Turcizmi u srpskohrvatskom jeziku“ Abdulaha Škaljića

Abdulah Škaljić’s dictionary Turcisms in Serbo-Croatian language con- tains 73 words that describe colours. However, our list records 82 lexemes that describe some of the colour nuances that Škaljić explains with a word, phrase or clause. The ex- cerpted material was classified on the basis of three criteria: light and dark colours, pale (whitish) and dark (blue) colours, and visual categorization of colours in painting (red and its shades, blue and its shades and yellow and its shades). The words describing red and the shades of red are the most frequent, while those related to yellow and the shades of yellow are the least frequent. Dark colours are prevalent. It has been noticed that the classification of colours could be based on some other criteria as well. That is why it is important to point out the need for this topic to remain open for future research studies, which would intend to figure out how and why so many words for colours disappeared from the Serbian language within just one century.