Mei-Chen Yeh

Bio: Mei-Chen Yeh is an academic researcher from National Taiwan Normal University. The author has contributed to research in topics: Contextual image classification & Video copy detection. The author has an hindex of 12, co-authored 49 publications receiving 2058 citations. Previous affiliations of Mei-Chen Yeh include University of California, Santa Barbara & Intel.

Fast Human Detection Using a Cascade of Histograms of Oriented Gradients

Abstract: We integrate the cascade-of-rejectors approach with the Histograms of Oriented Gradients (HoG) features to achieve a fast and accurate human detection system. The features used in our system are HoGs of variable-size blocks that capture salient features of humans automatically. Using AdaBoost for feature selection, we identify the appropriate set of blocks, from a large set of possible blocks. In our system, we use the integral image representation and a rejection cascade which significantly speed up the computation. For a 320 × 280 image, the system can process 5 to 30 frames per second depending on the density in which we scan the image, while maintaining an accuracy level similar to existing methods.

Video copy detection by fast sequence matching

Abstract: Sequence matching techniques are effective for comparing two videos. However, existing approaches suffer from demanding computational costs and thus are not scalable for large-scale applications. In this paper we view video copy detection as a local alignment problem between two frame sequences and propose a two-level filtration approach which achieves significant acceleration to the matching process. First, we propose to use an adaptive vocabulary tree to index all frame descriptors extracted from the video database. In this step, each video is treated as a "bag of frames." Such an indexing structure not only provides a rich vocabulary for representing videos, but also enables efficient computation of a pyramid matching kernel between videos. This vocabulary tree filters those videos that are dissimilar to the query based on their histogram pyramid representations. Second, we propose a fast edit-distance-based sequence matching method that avoids unnecessary comparisons between dissimilar frame pairs. This step reduces the quadratic runtime to a linear time with respect to the lengths of the sequences under comparison. Experiments on the MUSCLE VCD benchmark demonstrate that our approach is effective and efficient. It is 18X faster than the original sequence matching algorithms. This technique can be applied to several other visual retrieval tasks including shape retrieval. We demonstrate that the proposed method can also achieve a significant speedup for the shape retrieval task on the MPEG-7 shape dataset.

Multimodal fusion using learned text concepts for image categorization

Abstract: Conventional image categorization techniques primarily rely on low-level visual cues. In this paper, we describe a multimodal fusion scheme which improves the image classification accuracy by incorporating the information derived from the embedded texts detected in the image under classification. Specific to each image category, a text concept is first learned from a set of labeled texts in images of the target category using Multiple Instance Learning [1]. For an image under classification which contains multiple detected text lines, we calculate a weighted Euclidian distance between each text line and the learned text concept of the target category. Subsequently, the minimum distance, along with low-level visual cues, are jointly used as the features for SVM-based classification. Experiments on a challenging image database demonstrate that the proposed fusion framework achieves a higher accuracy than the state-of-art methods for image classification.

Image buffering techniques

Abstract: A system, apparatus, method and article to perform buffering techniques are described. The apparatus may include a buffer having a fixed number of storage slots that store reconstructed picture representations received from an image processing module. Also, the apparatus may include a buffer status unit to store a multiple information items to indicate one or more buffer characteristics of the buffer. Further, the apparatus may include a buffer control module to manage storage within the buffer.

A compact, effective descriptor for video copy detection

Abstract: Large scale video copy detection tasks require a compact and computational-efficient descriptor that is robust to various transformations that are typically applied to generate copies. In this paper, we propose a new frame-level descriptor for such a task. The descriptor encodes the internal structure of a video frame by computing the pair-wise correlations between geometrically pre-indexed blocks. It is conceptually simple, small in size, and fast to compute. Experiments using the MUSCLE VCD benchmark show its superior performance compared to existing approaches.

Pattern Recognition and Machine Learning

Abstract: Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks

Abstract: Face detection and alignment in unconstrained environment are challenging due to various poses, illuminations, and occlusions. Recent studies show that deep learning approaches can achieve impressive performance on these two tasks. In this letter, we propose a deep cascaded multitask framework that exploits the inherent correlation between detection and alignment to boost up their performance. In particular, our framework leverages a cascaded architecture with three stages of carefully designed deep convolutional networks to predict face and landmark location in a coarse-to-fine manner. In addition, we propose a new online hard sample mining strategy that further improves the performance in practice. Our method achieves superior accuracy over the state-of-the-art techniques on the challenging face detection dataset and benchmark and WIDER FACE benchmarks for face detection, and annotated facial landmarks in the wild benchmark for face alignment, while keeps real-time performance.

Pedestrian Detection: An Evaluation of the State of the Art

Abstract: Pedestrian detection is a key problem in computer vision, with several applications that have the potential to positively impact quality of life. In recent years, the number of approaches to detecting pedestrians in monocular images has grown steadily. However, multiple data sets and widely varying evaluation protocols are used, making direct comparisons difficult. To address these shortcomings, we perform an extensive evaluation of the state of the art in a unified framework. We make three primary contributions: 1) We put together a large, well-annotated, and realistic monocular pedestrian detection data set and study the statistics of the size, position, and occlusion patterns of pedestrians in urban scenes, 2) we propose a refined per-frame evaluation methodology that allows us to carry out probing and informative comparisons, including measuring performance in relation to scale and occlusion, and 3) we evaluate the performance of sixteen pretrained state-of-the-art detectors across six data sets. Our study allows us to assess the state of the art and provides a framework for gauging future efforts. Our experiments show that despite significant progress, performance still has much room for improvement. In particular, detection is disappointing at low resolutions and for partially occluded pedestrians.

Fast Feature Pyramids for Object Detection

Abstract: Multi-resolution image features may be approximated via extrapolation from nearby scales, rather than being computed explicitly. This fundamental insight allows us to design object detection algorithms that are as accurate, and considerably faster, than the state-of-the-art. The computational bottleneck of many modern detectors is the computation of features at every scale of a finely-sampled image pyramid. Our key insight is that one may compute finely sampled feature pyramids at a fraction of the cost, without sacrificing performance: for a broad family of features we find that features computed at octave-spaced scale intervals are sufficient to approximate features on a finely-sampled pyramid. Extrapolation is inexpensive as compared to direct feature computation. As a result, our approximation yields considerable speedups with negligible loss in detection accuracy. We modify three diverse visual recognition systems to use fast feature pyramids and show results on both pedestrian detection (measured on the Caltech, INRIA, TUD-Brussels and ETH data sets) and general object detection (measured on the PASCAL VOC). The approach is general and is widely applicable to vision algorithms requiring fine-grained multi-scale analysis. Our approximation is valid for images with broad spectra (most natural images) and fails for images with narrow band-pass spectra (e.g., periodic textures).

An HOG-LBP human detector with partial occlusion handling

Abstract: By combining Histograms of Oriented Gradients (HOG) and Local Binary Pattern (LBP) as the feature set, we propose a novel human detection approach capable of handling partial occlusion. Two kinds of detectors, i.e., global detector for whole scanning windows and part detectors for local regions, are learned from the training data using linear SVM. For each ambiguous scanning window, we construct an occlusion likelihood map by using the response of each block of the HOG feature to the global detector. The occlusion likelihood map is then segmented by Mean-shift approach. The segmented portion of the window with a majority of negative response is inferred as an occluded region. If partial occlusion is indicated with high likelihood in a certain scanning window, part detectors are applied on the unoccluded regions to achieve the final classification on the current scanning window. With the help of the augmented HOG-LBP feature and the global-part occlusion handling method, we achieve a detection rate of 91.3% with FPPW= 10−6, 94.7% with FPPW= 10−5, and 97.9% with FPPW= 10−4 on the INRIA dataset, which, to our best knowledge, is the best human detection performance on the INRIA dataset. The global-part occlusion handling method is further validated using synthesized occlusion data constructed from the INRIA and Pascal dataset.