Mark Barnard

Automatic analysis of multimodal group actions in meetings

This paper investigates the recognition of group actions in meetings. A framework is employed in which group actions result from the interactions of the individual participants. The group actions are modeled using different HMM-based approaches, where the observations are provided by a set of audiovisual features monitoring the actions of individuals. Experiments demonstrate the importance of taking interactions into account in modeling the group actions. It is also shown that the visual modality contains useful information, even for predominantly audio-based events, motivating a multimodal approach to meeting analysis.

Lipreading With Local Spatiotemporal Descriptors

Visual speech information plays an important role in lipreading under noisy conditions or for listeners with a hearing impairment. In this paper, we present local spatiotemporal descriptors to represent and recognize spoken isolated phrases based solely on visual input. Spatiotemporal local binary patterns extracted from mouth regions are used for describing isolated phrase sequences. In our experiments with 817 sequences from ten phrases and 20 speakers, promising accuracies of 62% and 70% were obtained in speaker-independent and speaker-dependent recognition, respectively. In comparison with other methods on AVLetters database, the accuracy, 62.8%, of our method clearly outperforms the others. Analysis of the confusion matrix for 26 English letters shows the good clustering characteristics of visemes for the proposed descriptors. The advantages of our approach include local processing and robustness to monotonic gray-scale changes. Moreover, no error prone segmentation of moving lips is needed.

Gradient field descriptor for sketch based retrieval and localization

We present an image retrieval system driven by free-hand sketched queries depicting shape. We introduce Gradient Field HoG (GF-HOG) as a depiction invariant image descriptor, encapsulating local spatial structure in the sketch and facilitating efficient codebook based retrieval. We show improved retrieval accuracy over 3 leading descriptors (Self Similarity, SIFT, HoG) across two datasets (Flickr160, ETHZ extended objects), and explain how GF-HOG can be combined with RANSAC to localize sketched objects within relevant images. We also demonstrate a prototype sketch driven photo montage application based on our system.

ℓ p norm multiple kernel Fisher discriminant analysis for object and image categorisation

In this paper, we generalise multiple kernel Fisher discriminant analysis (MK-FDA) such that the kernel weights can be regularised with an ℓp norm for any p ≥ 1, in contrast to existing MK-FDA that uses either l1 or l2 norm. We present formulations for both binary and multiclass cases and solve the associated optimisation problems efficiently with semi-infinite programming. We show on three object and image categorisation benchmarks that by learning the intrinsic sparsity of a given set of base kernels using a validation set, the proposed ℓp MK-FDA outperforms its fixed-norm counterparts, and is capable of producing state-of-the-art performance. Moreover, we show that our ℓp MK-FDA outperforms the ℓp multiple kernel support vector machine (ℓp MK-SVM) which has been recently proposed. Based on this observation and our experience with single kernel FDA and SVM, we argue that the almost century-old FDA is still a strong competitor of the popular SVM.

An evaluation of bags-of-words and spatio-temporal shapes for action recognition

Bags-of-visual-Words (BoW) and Spatio-Temporal Shapes (STS) are two very popular approaches for action recognition from video. The former (BoW) is an un-structured global representation of videos which is built using a large set of local features. The latter (STS) uses a single feature located on a region of interest (where the actor is) in the video. Despite the popularity of these methods, no comparison between them has been done. Also, given that BoW and STS differ intrinsically in terms of context inclusion and globality/locality of operation, an appropriate evaluation framework has to be designed carefully. This paper compares these two approaches using four different datasets with varied degree of space-time specificity of the actions and varied relevance of the contextual background. We use the same local feature extraction method and the same classifier for both approaches. Further to BoW and STS, we also evaluated novel variations of BoW constrained in time or space. We observe that the STS approach leads to better results in all datasets whose background is of little relevance to action classification.

Audio Assisted Robust Visual Tracking With Adaptive Particle Filtering

The problem of tracking multiple moving speakers in indoor environments has received much attention. Earlier techniques were based purely on a single modality, e.g., vision. Recently, the fusion of multi-modal information has been shown to be instrumental in improving tracking performance, as well as robustness in the case of challenging situations like occlusions (by the limited field of view of cameras or by other speakers). However, data fusion algorithms often suffer from noise corrupting the sensor measurements which cause non-negligible detection errors. Here, a novel approach to combining audio and visual data is proposed. We employ the direction of arrival angles of the audio sources to reshape the typical Gaussian noise distribution of particles in the propagation step and to weight the observation model in the measurement step. This approach is further improved by solving a typical problem associated with the PF, whose efficiency and accuracy usually depend on the number of particles and noise variance used in state estimation and particle propagation. Both parameters are specified beforehand and kept fixed in the regular PF implementation which makes the tracker unstable in practice. To address these problems, we design an algorithm which adapts both the number of particles and noise variance based on tracking error and the area occupied by the particles in the image. Experiments on the AV16.3 dataset show the advantage of our proposed methods over the baseline PF method and an existing adaptive PF algorithm for tracking occluded speakers with a significantly reduced number of particles.

Multi-modal audio-visual event recognition for football analysis

The recognition of events within multi-modal data is a challenging problem. In this paper we focus on the recognition of events by using both audio and video data. We investigate the use of data fusion techniques in order to recognise these sequences within the framework of hidden Markov models (HMM) used to model audio and video data sequences. Specifically we look at the recognition of play and break sequences in football and the segmentation of football games based on these two events. Recognising relatively simple semantic events such as this is an important step towards full automatic indexing of such video material. These experiments were done using approximately 3 hours of data from two games of the Euro96 competition. We propose that modelling the audio and video streams separately for each sequence and fusing the decisions from each stream should yield an accurate and robust method of segmenting multi-modal data.

Multimodal (Audio-Visual) Source Separation Exploiting Multi-Speaker Tracking, Robust Beamforming, and Time-Frequency Masking

A novel multimodal source separation approach is proposed for physically moving and stationary sources which exploits a circular microphone array, multiple video cameras, robust spatial beamforming and time-frequency masking. The challenge of separating moving sources, including higher reverberation time (RT) even for physically stationary sources, is that the mixing filters are time varying; as such the unmixing filters should also be time varying but these are difficult to determine from only audio measurements. Therefore in the proposed approach, visual modality is used to facilitate the separation for both stationary and moving sources. The movement of the sources is detected by a three-dimensional tracker based on a Markov Chain Monte Carlo particle filter. The audio separation is performed by a robust least squares frequency invariant data-independent beamformer. The uncertainties in source localisation and direction of arrival information obtained from the 3D video-based tracker are controlled by using a convex optimisation approach in the beamformer design. In the final stage, the separated audio sources are further enhanced by applying a binary time-frequency masking technique in the cepstral domain. Experimental results show that using the visual modality, the proposed algorithm cannot only achieve performance better than conventional frequency-domain source separations algorithms, but also provide acceptable separation performance for moving sources.

Robust playfield segmentation using MAP adaptation

A vital task in sports video annotation is to detect and segment areas of the playfield. This is an important first step in player or ball tracking and detecting the location of the play on the playfield. In this paper, we present a technique using statistical models, Gaussian mixture models (GMMs) and maximum a posteriori (MAP) adaptation. This involves first creating a generic model of the play-field colour and then using unsupervised MAP adaptation to adapt this model to the colour of the playfield in each game. This technique provides a robust and accurate segmentation of the playfield. To demonstrate the robustness of the method we tested it on a number of different sports that have grass playfields, rugby, soccer and field hockey.

Robust Multi-Speaker Tracking via Dictionary Learning and Identity Modeling

We investigate the problem of visual tracking of multiple human speakers in an office environment. In particular, we propose novel solutions to the following challenges: (1) robust and computationally efficient modeling and classification of the changing appearance of the speakers in a variety of different lighting conditions and camera resolutions; (2) dealing with full or partial occlusions when multiple speakers cross or come into very close proximity; (3) automatic initialization of the trackers, or re-initialization when the trackers have lost lock caused by e.g. the limited camera views. First, we develop new algorithms for appearance modeling of the moving speakers based on dictionary learning (DL), using an off-line training process. In the tracking phase, the histograms (coding coefficients) of the image patches derived from the learned dictionaries are used to generate the likelihood functions based on Support Vector Machine (SVM) classification. This likelihood function is then used in the measurement step of the classical particle filtering (PF) algorithm. To improve the computational efficiency of generating the histograms, a soft voting technique based on approximate Locality-constrained Soft Assignment (LcSA) is proposed to reduce the number of dictionary atoms (codewords) used for histogram encoding. Second, an adaptive identity model is proposed to track multiple speakers whilst dealing with occlusions. This model is updated online using Maximum a Posteriori (MAP) adaptation, where we control the adaptation rate using the spatial relationship between the subjects. Third, to enable automatic initialization of the visual trackers, we exploit audio information, the Direction of Arrival (DOA) angle, derived from microphone array recordings. Such information provides, a priori, the number of speakers and constrains the search space for the speakers faces. The proposed system is tested on a number of sequences from three publicly available and challenging data corpora (AV16.3, EPFL pedestrian data set and CLEAR) with up to five moving subjects.