Sampled Image Tagging and Retrieval Methods on User Generated Content

Abstract

Traditional image tagging and retrieval algorithms have limited value as a result of being trained with heavily curated datasets. These limitations are most evident when arbitrary search words are used that do not intersect with training set labels. Weak labels from user generated content (UGC) found in the wild (e.g., Google Photos, FlickR, etc.) have an almost unlimited number of unique words in the metadata tags. Prior work on word embeddings successfully leveraged unstructured text with large vocabularies, and our proposed method seeks to apply similar cost functions to open source imagery. Specifically, we train a deep learning image tagging and retrieval system on large-scale, user generated content (UGC) using sampling methods and joint optimization of word embeddings. By using the Yahoo! FlickR Creative Commons (YFCC100M) dataset, such an approach builds robustness to common unstructured data issues that include but are not limited to irrelevant tags, misspellings, multiple languages, polysemy, and tag imbalance. As a result, the final proposed algorithm will not only yield comparable results to state of the art in conventional image tagging, but will enable new capability to train algorithms on large, scale unstructured text in the YFCC100M dataset and outperform cited work in zero-shot capability.

From Intra-Modal to Inter-Modal Space: Multi-task Learning of Shared Representations for Cross-Modal Retrieval

“…Fast0Tag [14] projects an image by identifying a principal direction in the space and targeting that principal direction when learning to project the image. [15] uses noise contrastive estimation on a noisy web-scale dataset [16] to learn projection from image to word embeddings space. VSE++ [17] proposes a modified pairwise ranking loss weighted by violation caused by hard-negatives.…”