Track topics on Twitter Track topics that are important to you
Cross-domain shoe image retrieval is a challenging problem, because the query photo from the street domain (daily life scenario) and the reference photo in the online domain (online shop images) have significant visual differences due to the viewpoint and scale variation, self-occlusion, and cluttered background. This paper proposes the Semantic Hierarchy Of attributE Convolutional Neural Network (SHOE-CNN) with a three-level feature representation for discriminative shoe feature expression and efficient retrieval. The SHOE-CNN with its newly designed loss function systematically merges semantic attributes of closer visual appearances to prevent shoe images with the obvious visual differences being confused with each other; the features extracted from image, region, and part levels effectively match the shoe images across different domains. We collect a large-scale shoe dataset composed of 14341 street domain and 12652 corresponding online domain images with fine-grained attributes to train our network and evaluate our system. The top-20 retrieval accuracy improves significantly over the solution with the pre-trained CNN features.
This article was published in the following journal.
Name: IEEE transactions on image processing : a publication of the IEEE Signal Processing Society
A number of vision problems such as zero-shot learning and person re-identification can be considered as cross-class transfer learning problems. As mid-level semantic properties shared cross different...
A novel cross-domain image retrieval system that is based on a high-speed optical correlator with a coaxial holographic system is presented. Our newly designed conversion module for the optical correl...
Retrograde amnesia (RA), which includes loss of memory for past personal events (autobiographical RA) and for acquired knowledge (semantic RA), has been largely documented in patients with amnestic mi...
Shoe inserts and shoe modifications are used to reduce plantar peak pressure. The effects of different shoe inserts and shoe construction strategies for relief of the forefoot have not yet been suffic...
We study the problem of fine-grained sketch-based image retrieval. By performing instance-level (rather than category-level) retrieval, it embodies a timely and practical application, particularly wit...
Offloading is the concept of relieving pressure to help prevent/cure plantar ulcers, especially in diabetic patients. Many forms of offloading are available, with many limitations to apply...
This trial will assess the ability of shoe inserts to relieve pain and disability in people with knee osteoarthritis (OA) in the inner (medial) aspect of the knee.
Osteoarthritis is a degenerative joint disease and is the most common form of arthritis. This study will evaluate the effectiveness of customized shoe inserts in controlling and relieving ...
To evaluate the effect of wearing the 'Nike Free'Shoe on isokinetic muscle strength of foot and ankle and on proprioception the following protocol will be used. The test population consis...
This trial will test shoe inserts for the treatment of knee osteoarthritis, the most common form of knee arthritis. Those with disease on the inner (medial) aspect of the knee will be stud...
The perceived attribute of a sound which corresponds to the physical attribute of intensity.
A research and development program initiated by the NATIONAL LIBRARY OF MEDICINE to build knowledge sources for the purpose of aiding the development of systems that help health professionals retrieve and integrate biomedical information. The knowledge sources can be used to link disparate information systems to overcome retrieval problems caused by differences in terminology and the scattering of relevant information across many databases. The three knowledge sources are the Metathesaurus, the Semantic Network, and the Specialist Lexicon.
Studies which start with the identification of persons with a disease of interest and a control (comparison, referent) group without the disease. The relationship of an attribute to the disease is examined by comparing diseased and non-diseased persons with regard to the frequency or levels of the attribute in each group.
A family of transcription factors characterized by the presence of a bipartite DNA-binding domain known as the POU domain. The POU domain contains two subdomains, a POU-specific domain and a POU-homeodomain. The POU domain was originally identified as a region of approximately 150 amino acids shared between the Pit-1, Oct-1, Oct-2, and Unc-86 transcription factors.
A conserved protein interaction domain of the death domain superfamily that is structurally similar to the DEATH EFFECTOR DOMAIN and CASPASE RECRUITMENT DOMAIN. Death domains bind each other to form oligomers and occur on DEATH DOMAIN RECEPTORS, where they are required for APOPTOSIS signaling and non-apoptotic functions.