Landmark labeling of training images is essential for many learning tasks in computer vision, such as object detection, tracking, and alignment. Image labeling is typically conducted manually, which is both labor-intensive and error-prone. To improve this process, this paper proposes a new approach to estimate a set of landmarks for a large image ensemble with only a small number of manually labeled images from the ensemble. Our approach, named semi-supervised least-squares congealing, aims to minimize an objective function defined on both labeled and unlabeled images. A shape model is learnt on-line to constrain the landmark configuration. We also employ a partitioning strategy to allow coarse-to-fine landmark estimation. Extensive experiments on facial images show that our approach can reliably and accurately label landmarks for a large image ensemble starting from a small number of manually labeled images, under various challenging scenarios.

Joint alignment for an image ensemble can rectify images in the spatial domain such that the aligned images are as similar to each other as possible. This important technology has been applied to various object classes and medical applications. However, previous approaches to joint alignment work on an ensemble of a single object class. Given an ensemble with multiple object classes, we propose an approach to automatically and simultaneously solve two problems, image alignment and clustering. Both the alignment parameters and clustering parameters are formulated into a unified objective function, whose optimization leads to an unsupervised joint estimation approach. It is further extended to semi-supervised simultaneous estimation where a few labeled images are provided. Extensive experiments on diverse real-world databases demonstrate the capabilities of our work on this challenging problem.

Active Appearance Models (AAMs) represent the shape and appearance of an object via two low-dimensional subspaces, one for shape and one for appearance. AAMs for facial images are currently receiving considerable attention from the computer vision community. However, most existing work focuses on fitting AAMs to a single image. For many applications, effectively fitting an AAM to video sequences is of critical importance and challenging, especially considering the varying quality of real-world video content. This paper proposes a hybrid model to address this problem. Both a generic AAM and a subject-specific model are employed simultaneously in the proposed fitting scheme. Experimental results from outdoor surveillance video sequences demonstrate the improved image registration across video frames and faster fitting convergence. 1

Active Appearance Models (AAMs) represent the shape and appearance of an object via two low-dimensional subspaces, one for shape and one for appearance. AAMs for facial images are currently receiving considerable attention from the vision community. However, most existing work focuses on fitting AAMs to high-quality facial images. For many applications, effectively fitting an AAM to low-resolution facial images is of critical importance. This paper addresses this challenge from two aspects. On the modeling side, we propose an iterative AAM enhancement scheme, which not only results in increased fitting speed, but also improves the fitting robustness. For fitting AAMs to low-resolution images, we build a multi-resolution AAM and show that the best fitting performance is obtained when the model resolution is slightly higher than the facial image resolution. Experimental results using both indoor video and outdoor surveillance video are presented. 1

A variant of the SPIHT image compression algorithm called no list SPIHT (NLS) is presented. NLS operates without linked lists and is suitable for a fast, simple hardware implementation. NLS has a fixed predetermined memory requirement about 50% larger than that needed for the image alone. Instead of lists, a state table with four bits per coefficient keeps track of the set partitions and what information has been encoded. NLS sparsely marks selected descendant nodes of insignificant trees in the state table in such a way that large groups of predictably insignificant pixels are easily identified and skipped during coding passes. The image data is stored in a one dimensional recursive zig-zag array for computational efficiency and algorithmic simplicity. The performance of the algorithm on standard test images is nearly the same as SPIHT.

This paper describes a low-memory cache efficient hybrid block coder (HBC) for images in which an image subband decomposition is partitioned into a combination of spatial blocks and subband blocks, which are independently coded. Spatial blocks contain hierarchical trees spanning subband levels, and are each encoded using the SPIHT algorithm. Subband blocks contain a block of coefficients from within a single subband, and are each encoded by the SPECK algorithm. The decomposition may have the dyadic or a wavelet packet structure. Rate is allocated amongst the sub-bitstreams produced for each block and they are packetized. The partitioning structure supports resolution embedding. The final bitstream may be progressive in fidelity or in resolution.