Dynamically Visual Disambiguation of Keyword based Image Search

By Published On: January 4, 20210.2 min readViews: 247

Due to the high cost of manual annotation, learning directly from the web has attracted broad attention. One issue that limits their performance is the problem of visual polysemy. To address this issue, we present an adaptive multi-model framework that resolves polysemy by visual disambiguation.

Download PDF

Related Research Papers