A drone monitoring system that integrates deeplearning-based detection and tracking modules is proposed in this work. The biggest challenge in adopting deep learning methods for drone detection is the limited amount of training drone images. To address this issue, we develop a model-based drone augmentation technique that automatically generates drone images with a bounding box label on drone's location. To track a small flying drone, we utilize the residual information between consecutive image frames. Finally, we present an integrated detection and tracking system that outperforms the performance of each individual module containing detection or tracking only. The experiments show that, even being trained on synthetic data, the proposed system performs well on real world drone images with complex background. The USC drone detection and tracking dataset with user labeled bounding boxes is available to the public.

Summary Opportunistic Networks (OppNets) are intermittently connected infrastructure less wireless networks. There is no continuous end‐to‐end connection between the sender and the receiver, and hence nodes follow a store‐carry‐forward mechanism. The routing algorithm is required to be adaptive to the changing topology of the network. In this work, Q‐Routing technique has been used with forwarding probability and incorporated using Poisson's probability for decision making and controlling transmission energy. The algorithm refines the forwarding decision of finding the next suitable hop by exploiting the characteristics of nodes such as daily routines, mobility pattern, etc. In simulations, the performance of PBQ‐Routing is compared with Q‐Routing, Epidemic Routing, PRoPHET (Probabilistic Routing Protocol using History of Encounters and Transitivity), and HBPR (History Based Prediction Routing) for OppNets. The use of Poisson's distribution improves the effectiveness of the probabilistic forwarding decision. The findings show that the delivery probability of PBQ‐Routing almost gets doubled and overhead ratio reduces to half in comparison with that of Q‐Routing when used in OppNets. PBQ‐Routing outperforms other Q‐Routing based algorithms over Poisson's distribution. As there is less replication in case of PBQ‐Routing, it also saves the transmission energy.

Text-visual (or called semantic-visual) embedding is a central problem in vision-language research. It typically involves mapping of an image and a text description to a common feature space through a CNN image encoder and a RNN language encoder. In this paper, we propose a new method for learning text-visual embedding using both image titles and click-through data from an image search engine. We also propose a new triplet loss function by modeling positive awareness of the embedding, and introduce a novel mini-batch-based hard negative sampling approach for better data efficiency in the learning process. Experimental results show that our proposed method outperforms existing methods, and is also effective for real-world textto-visual retrieval.

We present StyleBabel, a unique open access dataset of natural language captions and free-form tags describing the artistic style of over 135K digital artworks, collected via a novel participatory method from experts studying at specialist art and design schools. StyleBabel was collected via an iterative method, inspired by 'Grounded Theory': a qualitative approach that enables annotation while co-evolving a shared language for fine-grained artistic style attribute description. We demonstrate several downstream tasks for StyleBabel, adapting the recent ALADIN architecture for fine-grained style similarity, to train cross-modal embeddings for: 1) free-form tag generation; 2) natural language description of artistic style; 3) fine-grained text search of style. To do so, we extend ALADIN with recent advances in Visual Transformer (ViT) and cross-modal representation learning, achieving a state of the art accuracy in fine-grained style retrieval.

No abstract