Takayuki Kurozumi

A fast template matching algorithm with adaptive skipping using inner-subtemplates' distances

This paper proposes a new fast template matching algorithm that skips comparison between a template and search windows neighboring an already compared dissimilar sub-window. Comparison skipping is executed when a lower bound of distance between the template and a window exceeds a threshold. The lower bound of distance between the template and the window is determined by the triangular inequality in distances: the distance between a subtemplate and a subwindow and that between inter-subtemplates. Experimental results demonstrate that the proposed method is up to five times faster than the conventional fast exhaustive search method (sequential similarity detection algorithm), strictly guaranteeing the same accuracy.

A Quick Search Method for Audio Signals Based on a Piecewise Linear Representation of Feature Trajectories

This paper presents a new method for a quick similarity-based search through long unlabeled audio streams to detect and locate audio clips provided by users. The method involves feature-dimension reduction based on a piecewise linear representation of a sequential feature trajectory extracted from a long audio stream. Two techniques enable us to obtain a piecewise linear representation: the dynamic segmentation of feature trajectories and the segment-based Karhunen-Loeve (KL) transform. The proposed search method guarantees the same search results as the search method without the proposed feature-dimension reduction method in principle. Experimental results indicate significant improvements in search speed. For example, the proposed method reduced the total search time to approximately 1/12 that of previous methods and detected queries in approximately 0.3 s from a 200-h audio database.

Robust Search Methods for Music Signals Based on Simple Representation

Signal similarity search is an important technique for music information retrieval. A basic task is finding identical signal segments on unlabeled music-signal archives, given a short music signal fragment as a query. In such a task, the search must be fast and sufficiently robust against possible signal fluctuations due to noise and distortions. In this special session paper, we describe a search method designed to cope with additive interfering sounds by spectral partitioning. Then, we introduce another method designed to be robust under multiplicative noise or distortion based on binary area representation.

A method for robust and quick video searching using probabilistic dither-voting

We propose a quick and accurate search method for detecting a query signal from long video recordings The method is based on the time-series active search, which is a quick searching method for audio and video signals that we previously proposed. Time-series active search is based on a histogram matching scheme and an efficient pruning mechanism, and therefore, it was very quick. We found, however, that the accuracy sometimes deteriorates when it is applied to searches through long video archives that are composed of many similar video images or those containing feature distortions caused by video dubbing or low-bit-rate compression. The problem arises from (1) insufficient capability of representing features and (2) feature distortions. Thus, the method proposed here uses LBG-based VQ to improve the capacity to represent features and probabilistic dither-voting to improve robustness with respect to feature distortions. The experiments prove the effects of the proposed method.

A quick search method for multimedia signals using feature compression based on piecewise linear maps

We propose a quick algorithm for multimedia signal search. The algorithm comprises two techniques: feature compression based on piecewise linear maps and distance bounding to efficiently limit the search space. When compared with existing multimedia search techniques, they greatly reduce the computational cost required in searching. Although feature compression is employed in our method, our bounding technique mathematically guarantees the same recall rate as the search based on the original features; no segment to be detected is missed. Experiments indicate that the proposed algorithm is approximately 10 times faster than and as accurate as an existing fast method maitaining the same search accuracy.

A quick video search method based on local and global feature clustering

This paper proposes a quick method of similarity-based video searching to detect and locate a specific video clip given as a query in a stored long video stream. The method employs a two-stage process: local and global feature clustering. The local clustering exploits continuity or local similarities between video features, and the global clustering gathers similar video frames that are not necessarily adjacent to each other. These processes prune irrelevant sections on a video stream. The method guarantees the exactly same search result as the exhaustive search. Experiments performed on a PC show that the proposed method can correctly detect and locate a 7.5-second clip in a 150-hour video recording in 15 ms on average.

Feature fluctuation absorption for a quick audio retrieval from long recordings

Kashino et al. proposed (1999) a histogram-based quick signal search method called time-series active search (TAS). TAS has only been effective in the exact matching case, where the segments to be detected are assumed to be exactly same as the reference signal. Here, we extend the method so that it is applicable even if the features fluctuate. In addition to the feature modification, feature dithering is discussed to absorb feature fluctuations. Efficient time-scaled search is also investigated to cope with variations of the reference signal duration. Tests using broadcast recordings show that the extended method improves the accuracy in nonexact-matching tasks such as hand-clap detection and word spotting in a single-speakers narration. The tests also show the speed-ups by pruning introduced in the time-scaled search.

Dynamic-segmentation-based feature dimension reduction for quick audio/video searching

We propose a new feature dimension reduction method for multimedia search. The main technique in the method is dynamic segmentation that partitions sequential feature trajectories dynamically. While dynamic segmentation reduces the average dimensionality and accelerates the search, it requires huge amount of calculation. Thus, our method quickly executes suboptimal partitioning of the trajectories by using the discreteness of dimension changes. This guarantees the optimal amount of calculation to derive the suboptimal partitioning under the condition that the dimension monotonously increases as the segment length increases. The experiment shows that our method is over 10 times faster than a straightforward dynamic segmentation method.

A quick search method for multimedia signals using global pruning

The authors propose a new method for quickly searching for a specific audio or video signal to be detected within a long, stored audio or video stream to determine segments that contain signals that are nearly identical to the given signal. The Time-series Active Search (TAS) method is one of the quick search methods that have been proposed previously. This signal searching technique based on histograms extracted from the signals had implemented quick searching by local pruning, that is, omitting comparisons of segments for which searching was unnecessary based on similarities in the vicinity of the matching window. In contrast, the proposed technique implements significantly quicker searching by introducing global pruning, which looks at the entire signal time-series according to histogram classifications based on similarities of the entire signal to eliminate segments that need not be searched, in addition to local pruning. In this paper, the authors present a detailed discussion of the relationship between the degree of global pruning and the accuracy that is guaranteed. For example, the authors showed through experiments that when 128-dimension histograms were classified to 1024 clusters, the proposed technique achieved a search speed approximately 9 times that of TAS while preserving the same degree of accuracy. The preprocessing calculation time increased by approximately 1% of the time for playing the signal.

A fast audio search method based on skipping irrelevant signals by similarity upper-bound calculation

In this paper, we describe an approach to accelerate fingerprint techniques by skipping the search for irrelevant sections of the signal and demonstrate its application to the divide and locate (DAL) audio fingerprint method. The search result for the applied method, DAL3, is the same as that of DAL mathematically. Experimental results show that DAL3 can reduce the computational cost of DAL to approximately 25% for the task of music signal retrieval.