English

Semantically-Aware Attentive Neural Embeddings for Image-based Visual Localization

Computer Vision and Pattern Recognition 2019-07-04 v2

Abstract

We present an approach that combines appearance and semantic information for 2D image-based localization (2D-VL) across large perceptual changes and time lags. Compared to appearance features, the semantic layout of a scene is generally more invariant to appearance variations. We use this intuition and propose a novel end-to-end deep attention-based framework that utilizes multimodal cues to generate robust embeddings for 2D-VL. The proposed attention module predicts a shared channel attention and modality-specific spatial attentions to guide the embeddings to focus on more reliable image regions. We evaluate our model against state-of-the-art (SOTA) methods on three challenging localization datasets. We report an average (absolute) improvement of 19%19\% over current SOTA for 2D-VL. Furthermore, we present an extensive study demonstrating the contribution of each component of our model, showing 88--15%15\% and 4%4\% improvement from adding semantic information and our proposed attention module. We finally show the predicted attention maps to offer useful insights into our model.

Keywords

Cite

@article{arxiv.1812.03402,
  title  = {Semantically-Aware Attentive Neural Embeddings for Image-based Visual Localization},
  author = {Zachary Seymour and Karan Sikka and Han-Pang Chiu and Supun Samarasekera and Rakesh Kumar},
  journal= {arXiv preprint arXiv:1812.03402},
  year   = {2019}
}

Comments

Appearing in BMVC 2019

R2 v1 2026-06-23T06:36:25.312Z