This paper addresses the problem of unsupervised learning binary hash codes for efficient cross-modal retrieval. Many unimodal hashing studies have proven that both similarity preservation data and maintenance quantization quality are essential improving retrieval performance with codes. However, most existing methods mainly focused on former, latter still remains almost untouched. We propose a method to minimize errors, which is tailored hashing. Our approach, named Alternating...

This paper studies clothing and attribute recognition in the fashion domain. Specifically, this paper, we turn our attention to compatibility of items attributes (Fig 1). For example, people do not wear a skirt dress at same time, yet jacket shirt are preferred combination. We consider such inter-object or inter-attribute formulate Conditional Random Field (CRF) that seeks most probable combination given picture. The model takes into account location-specific appearance with respect human...

Article Free Access Share on PanoramaExcerpts: extracting and packing panoramas for video browsing Authors: Yukinobu Taniguchi NTT Human Interface Laboratories, Nippon Telegraph Telephone Corporation, 1-1 Hikari-no-oka, Yokosuka-Shi, Kanagawa, 239 Japan JapanView Profile , Akihito Akutsu Yoshinobu Tonomura Authors Info & Claims MULTIMEDIA '97: Proceedings of the fifth ACM international conference MultimediaNovember 1997 Pages 427–436https://doi.org/10.1145/266180.266396Online:01 November...

No abstract available.

Estimating the nutritional value of food based on image recognition is important to health support services employing mobile devices. The estimation accuracy can be improved by recognizing regions objects and ingredients contained in those regions. In this paper, we propose a method that estimates information segmentation labeling an adopting semantic method, which consider recipes as corresponding sets images ingredient labels. Any object or test annotated long training image, even if menu...

Many studies on action recognition from the third-person viewpoint have shown that articulated human pose can directly describe motion and is invariant to view change. However, conventional algorithms estimate cannot handle ego-centric images because they assume whole figure appears in image; only a few parts of body appear images. In this paper, we propose novel method for RGB-D Our extract by integrating hand detection, camera estimation, time-series filtering with constraint shape....

This paper proposes a method for the multi-camera tracking of dairy cows using rotated bounding boxes instead aligned to horizontal and vertical axes cameras. Experiments conducted on videos taken in an actual barn demonstrate effectiveness proposal.

A novel region-based approach is proposed to model semantic concepts using web images. Web images are mined obtain multiple visual patterns automatically that then used a concept. First, the salient region groups corresponding representative of concept and selected as positive samples. Next, pattern built in each group by BDA classifier. Finally all aggregated describe ensemble approach. Because method models utilizing patterns, it enhances variability when learning from diverse improves...

Content-based recommendation is a popular framework for video recommendation, where the videos recommended are selected according to content similarity. Aiming at providing semantically similar those already viewed by user, most existing methods measure similarity from tags or semantics-oriented features of videos. However, effective recommendations can also be based on affective content, which might more significantly correlated users' tastes and moods. We propose combine semantic...

We propose a practical method that realizes radiometric compensation based on direct and indirect light transports so uncalibrated projector-camera systems can create desired image displays 3D objects. By introducing compressed sensing, we efficiently construct the transport matrix includes direct/indirect reflected from unknown Assuming each camera pixel receives linear summation of reflection lights via color-mixing matrices, which are reformed by reference to transports, our proposed...

Abstract We find that pseudo‐stereoanomalous observers, who did not use disparity information for depth judgments in 3D‐graphics environments although they were stereoanomalous, tended to be younger (21.4%, 17–24 years old) than young‐middle age (11.4%, 25–39 old). Approximately 80% of older (50–83 also observers.

Digital signage (DS) is becoming well accepted as a ubiquitous medium that can deliver and display various types of information by specifying the location time. In this paper, we put our focus on following two directions DS development: (1) Networked DS: networking displays to form large-scale open network allows content providers submit contents, schedule, deliver, contents at right place time, (2) Context-aware displaying appropriate current status people, places, objects around display,...

Japanese comics (called manga) are traditionally created in monochrome format. In recent years, addition to comics, full color a more attractive medium, have appeared. Unfortunately, require manual colorization, which incurs high labor costs. Although automatic colorization methods been recently proposed, most of them designed for illustrations, not comics. Unlike since composed many consecutive images, the painting style must be consistent. To realize consistent we propose here...

This paper proposes a novel geometric verification method to handle 3D viewpoint changes under cluttered scenes for robust object recognition. Since previous voting-based approaches, which enable recognition in scenes, are based on 2D affine transformation, accuracy is significantly degraded when occur objects that abound real-world scenes. The view-directional consistency constraints requires the angle between observed directions of all matched feature points two given images must be...

A bus transportation system provides social and economic advantages plays a key role in our daily lives. This study aims at utilizing videos recorded by cameras installed buses to improve optimize services, where the required basic important information is passenger flow. We here consider problem of estimating flow as person re-identification on bus. Then, we propose method for improving identification accuracy persons using bus-specific properties. Results show that proposed simple but...

We propose a practical method that efficiently acquires light transports from arbitrary objects by using push broom-type projector illumination; vertical-strip and horizontal-strip patterns are used. The combinations of camera responses allow our to obtain between each overlapped pixel the response detected in common. Our significantly reduces number images needed for acquiring all compared brute-force scanning. Tests proposed real show it obtains more accurate various than scanning,...

At present, manhole covers seem to cover a large percentage of the road surface. If becomes worn, it must be replaced because worn can cause skidding and accidents. Staff have check condition all periodically, replace them if they are damaged. Because takes too much time manpower manually, system that automatically by some remote sensor is urgently required. However, detection wear has not been well studied. Our solution use image processing. We propose example-based approach using LBP...