Chenliang Xu

Streaming hierarchical video segmentation

The use of video segmentation as an early processing step in video analysis lags behind the use of image segmentation for image analysis, despite many available video segmentation methods. A major reason for this lag is simply that videos are an order of magnitude bigger than images; yet most methods require all voxels in the video to be loaded into memory, which is clearly prohibitive for even medium length videos. We address this limitation by proposing an approximation framework for streaming hierarchical video segmentation motivated by data stream algorithms: each video frame is processed only once and does not change the segmentation of previous frames. We implement the graph-based hierarchical segmentation method within our streaming framework; our method is the first streaming hierarchical video segmentation method proposed. We perform thorough experimental analysis on a benchmark video data set and longer videos. Our results indicate the graph-based streaming hierarchical method outperforms other streaming video segmentation methods and performs nearly as well as the full-video hierarchical graph-based method.

Evaluation of super-voxel methods for early video processing

Supervoxel segmentation has strong potential to be incorporated into early video analysis as superpixel segmentation has in image analysis. However, there are many plausible supervoxel methods and little understanding as to when and where each is most appropriate. Indeed, we are not aware of a single comparative study on supervoxel segmentation. To that end, we study five supervoxel algorithms in the context of what we consider to be a good supervoxel: namely, spatiotemporal uniformity, object/region boundary detection, region compression and parsimony. For the evaluation we propose a comprehensive suite of 3D volumetric quality metrics to measure these desirable supervoxel characteristics. We use three benchmark video data sets with a variety of content-types and varying amounts of human annotations. Our findings have led us to conclusive evidence that the hierarchical graph-based and segmentation by weighted aggregation methods perform best and almost equally-well on nearly all the metrics and are the methods of choice given our proposed assumptions.

A Thousand Frames in Just a Few Words: Lingual Description of Videos through Latent Topics and Sparse Object Stitching

The problem of describing images through natural language has gained importance in the computer vision community. Solutions to image description have either focused on a top-down approach of generating language through combinations of object detections and language models or bottom-up propagation of keyword tags from training images to test images through probabilistic or nearest neighbor techniques. In contrast, describing videos with natural language is a less studied problem. In this paper, we combine ideas from the bottom-up and top-down approaches to image description and propose a method for video description that captures the most relevant contents of a video in a natural language description. We propose a hybrid system consisting of a low level multimodal latent topic model for initial keyword annotation, a middle level of concept detectors and a high level module to produce final lingual descriptions. We compare the results of our system to human descriptions in both short and long forms on two datasets, and demonstrate that final system output has greater agreement with the human descriptions than any single level.

Flattening Supervoxel Hierarchies by the Uniform Entropy Slice

Supervoxel hierarchies provide a rich multiscale decomposition of a given video suitable for subsequent processing in video analysis. The hierarchies are typically computed by an unsupervised process that is susceptible to under-segmentation at coarse levels and over-segmentation at fine levels, which make it a challenge to adopt the hierarchies for later use. In this paper, we propose the first method to overcome this limitation and flatten the hierarchy into a single segmentation. Our method, called the uniform entropy slice, seeks a selection of supervoxels that balances the relative level of information in the selected supervoxels based on some post hoc feature criterion such as object-ness. For example, with this criterion, in regions nearby objects, our method prefers finer supervoxels to capture the local details, but in regions away from any objects we prefer coarser supervoxels. We formulate the uniform entropy slice as a binary quadratic program and implement four different feature criteria, both unsupervised and supervised, to drive the flattening. Although we apply it only to supervoxel hierarchies in this paper, our method is generally applicable to segmentation tree hierarchies. Our experiments demonstrate both strong qualitative performance and superior quantitative performance to state of the art baselines on benchmark internet videos.

LIBSVX: A Supervoxel Library and Benchmark for Early Video Processing

Supervoxel segmentation has strong potential to be incorporated into early video analysis as superpixel segmentation has in image analysis. However, there are many plausible supervoxel methods and little understanding as to when and where each is most appropriate. Indeed, we are not aware of a single comparative study on supervoxel segmentation. To that end, we study seven supervoxel algorithms, including both off-line and streaming methods, in the context of what we consider to be a good supervoxel: namely, spatiotemporal uniformity, object/region boundary detection, region compression and parsimony. For the evaluation we propose a comprehensive suite of seven quality metrics to measure these desirable supervoxel characteristics. In addition, we evaluate the methods in a supervoxel classification task as a proxy for subsequent high-level uses of the supervoxels in video analysis. We use six existing benchmark video datasets with a variety of content-types and dense human annotations. Our findings have led us to conclusive evidence that the hierarchical graph-based (GBH), segmentation by weighted aggregation (SWA) and temporal superpixels (TSP) methods are the top-performers among the seven methods. They all perform well in terms of segmentation accuracy, but vary in regard to the other desiderata: GBH captures object boundaries best; SWA has the best potential for region compression; and TSP achieves the best undersegmentation error.

Can humans fly? Action understanding with multiple classes of actors

Can humans fly? Emphatically no. Can cars eat? Again, absolutely not. Yet, these absurd inferences result from the current disregard for particular types of actors in action understanding. There is no work we know of on simultaneously inferring actors and actions in the video, not to mention a dataset to experiment with. Our paper hence marks the first effort in the computer vision community to jointly consider various types of actors undergoing various actions. To start with the problem, we collect a dataset of 3782 videos from YouTube and label both pixel-level actors and actions in each video. We formulate the general actor-action understanding problem and instantiate it at various granularities: both video-level single- and multiple-label actor-action recognition and pixel-level actor-action semantic segmentation. Our experiments demonstrate that inference jointly over actors and actions outperforms inference independently over them, and hence concludes our argument of the value of explicit consideration of various actors in comprehensive action understanding.

Deep Cross-Modal Audio-Visual Generation

Cross-modal audio-visual perception has been a long-lasting topic in psychology and neurology, and various studies have discovered strong correlations in human perception of auditory and visual stimuli. Despite work on computational multimodal modeling, the problem of cross-modal audio-visual generation has not been systematically studied in the literature. In this paper, we make the first attempt to solve this cross-modal generation problem leveraging the power of deep generative adversarial training. Specifically, we use conditional generative adversarial networks to achieve cross-modal audio-visual generation of musical performances. We explore different encoding methods for audio and visual signals, and work on two scenarios: instrument-oriented generation and pose-oriented generation. Being the first to explore this new problem, we compose two new datasets with pairs of images and sounds of musical performances of different instruments. Our experiments using both classification and human evaluation demonstrate that our model has the ability to generate one modality, i.e., audio/visual, from the other modality, i.e., visual/audio, to a good extent. Our experiments on various design choices along with the datasets will facilitate future research in this new problem space.

Actor-Action Semantic Segmentation with Grouping Process Models

Actor-action semantic segmentation made an important step toward advanced video understanding: what action is happening, who is performing the action, and where is the action happening in space-time. Current methods based on layered CRFs for this problem are local and unable to capture the long-ranging interactions of video parts. We propose a new model that combines the labeling CRF with a supervoxel hierarchy, where supervoxels at various scales provide cues for possible groupings of nodes in the CRF to encourage adaptive and long-ranging interactions. The new model defines a dynamic and continuous process of information exchange: the CRF influences what supervoxels in the hierarchy are active, and these active supervoxels, in turn, affect the connectivities in the CRF, we hence call it a grouping process model. By further incorporating the video-level recognition, the proposed method achieves a large margin of 60% relative improvement over the state of the art on the recent A2D large-scale video labeling dataset, which demonstrates the effectiveness of our modeling.

Dancelets Mining for Video Recommendation Based on Dance Styles

Dance is a unique and meaningful type of human expression, composed of abundant and various action elements. However, existing methods based on associated texts and spatial visual features have difficulty capturing the highly articulated motion patterns. To overcome this limitation, we propose to take advantage of the intrinsic motion information in dance videos to solve the video recommendation problem. We present a novel system that recommends dance videos based on a mid-level action representation, termed Dancelets. The Dancelets are used to bridge the semantic gap between video content and high-level concept, dance style, which plays a significant role in characterizing different types of dances. The proposed method executes automatic mining of dancelets with a concatenation of normalized cut clustering and linear discriminant analysis. This ensures that the discovered dancelets are both representative and discriminative. Additionally, to exploit the motion cues in videos, we employ motion boundaries as saliency priors to generate volumes of interest and extract C3D features to capture spatiotemporal information from the mid-level patches. Extensive experiments validated on our proposed large dance dataset, HIT Dances dataset, demonstrate the effectiveness of the proposed methods for dance style-based video recommendation.

Weakly Supervised Actor-Action Segmentation via Robust Multi-task Ranking

Fine-grained activity understanding in videos has attracted considerable recent attention with a shift from action classification to detailed actor and action understanding that provides compelling results for perceptual needs of cutting-edge autonomous systems. However, current methods for detailed understanding of actor and action have significant limitations: they require large amounts of finely labeled data, and they fail to capture any internal relationship among actors and actions. To address these issues, in this paper, we propose a novel, robust multi-task ranking model for weakly supervised actor-action segmentation where only video-level tags are given for training samples. Our model is able to share useful information among different actors and actions while learning a ranking matrix to select representative supervoxels for actors and actions respectively. Final segmentation results are generated by a conditional random field that considers various ranking scores for different video parts. Extensive experimental results on the Actor-Action Dataset (A2D) demonstrate that the proposed approach outperforms the state-of-the-art weakly supervised methods and performs as well as the top-performing fully supervised method.