ICCV 2015 论文列表

Visual object tracking is challenging as target objects often undergo significant appearance changes caused by deformation, abrupt motion, background clutter and occlusion. In this paper, we exploit features extracted from deep convolutional neural networks trained on object recognition datasets to improve tracking accuracy and robustness. The outputs of the last convolutional layers encode the semantic information of targets and such representations are robust to significant appearance variations. However, their spatial resolution is too coarse to precisely localize targets. In contrast, earlier convolutional layers provide more precise localization but are less invariant to appearance changes. We interpret the hierarchies of convolutional layers as a nonlinear counterpart of an image pyramid representation and exploit these multiple levels of abstraction for visual tracking. Specifically, we adaptively learn correlation filters on each convolutional layer to encode the target appearance. We hierarchically infer the maximum response of each layer to locate targets. Extensive experimental results on a largescale benchmark dataset show that the proposed algorithm performs favorably against state-of-the-art methods.

Recent work on scene classification still makes use of generic CNN features in a rudimentary manner. In this paper, we present a novel pipeline built upon deep CNN features to harvest discriminative visual objects and parts for scene classification. We first use a region proposal technique to generate a set of high-quality patches potentially containing objects, and apply a pre-trained CNN to extract generic deep features from these patches. Then we perform both unsupervised and weakly supervised learning to screen these patches and discover discriminative ones representing category-specific objects and parts. We further apply discriminative clustering enhanced with local CNN fine-tuning to aggregate similar objects and parts into groups, called meta objects. A scene image representation is constructed by pooling the feature response maps of all the learned meta objects at multiple spatial scales. We have confirmed that the scene image representation obtained using this new pipeline is capable of delivering state-of-the-art performance on two popular scene benchmark datasets, MIT Indoor 67 [22] and Sun397 [31].

We introduce a new benchmark "Humans Interacting with Common Objects" (HICO) for recognizing human-object interactions (HOI). We demonstrate the key features of HICO: a diverse set of interactions with common object categories, a list of well-defined, sense-based HOI categories, and an exhaustive labeling of co-occurring interactions with an object category in each image. We perform an in-depth analysis of representative current approaches and show that DNNs enjoy a significant edge. In addition, we show that semantic knowledge can significantly improve HOI recognition, especially for uncommon categories.

In image classification, visual separability between different object categories is highly uneven, and some categories are more difficult to distinguish than others. Such difficult categories demand more dedicated classifiers. However, existing deep convolutional neural networks (CNN) are trained as flat N-way classifiers, and few efforts have been made to leverage the hierarchical structure of categories. In this paper, we introduce hierarchical deep CNNs (HD-CNNs) by embedding deep CNNs into a category hierarchy. An HD-CNN separates easy classes using a coarse category classifier while distinguishing difficult classes using fine category classifiers. During HD-CNN training, component-wise pretraining is followed by global finetuning with a multinomial logistic loss regularized by a coarse category consistency term. In addition, conditional executions of fine category classifiers and layer parameter compression make HD-CNNs scalable for large-scale visual recognition. We achieve state-of-the-art results on both CIFAR100 and large-scale ImageNet 1000-class benchmark datasets. In our experiments, we build up three different HD-CNNs and they lower the top-1 error of the standard CNNs by 2.65%, 3.1% and 1.1%, respectively

The state-of-the-art salient object detection models are able to perform well for relatively simple scenes, yet for more complex ones, they still have difficulties in highlighting salient objects completely from background, largely due to the lack of sufficiently robust features for saliency prediction. To address such an issue, this paper proposes a novel hierarchy-associated feature construction framework for salient object detection, which is based on integrating elementary features from multi-level regions in a hierarchy. Furthermore, multi-layered deep learning features are introduced and incorporated as elementary features into this framework through a compact integration scheme. This leads to a rich feature representation, which is able to represent the context of the whole object/background and is much more discriminative as well as robust for salient object detection. Extensive experiments on the most widely used and challenging benchmark datasets demonstrate that the proposed approach substantially outperforms the state-of-the-art on salient object detection.

In this work we focus on the problem of image caption generation. We propose an extension of the long short term memory (LSTM) model, which we coin gLSTM for short. In particular, we add semantic information extracted from the image as extra input to each unit of the LSTM block, with the aim of guiding the model towards solutions that are more tightly coupled to the image content. Additionally, we explore different length normalization strategies for beam search to avoid bias towards short sentences. On various benchmark datasets such as Flickr8K, Flickr30K and MS COCO, we obtain results that are on par with or better than the current state-of-the-art.

Rotation search has become a core routine for solving many computer vision problems. The aim is to rotationally align two input point sets with correspondences. Recently, there is significant interest in developing globally optimal rotation search algorithms. A notable weakness of global algorithms, however, is their relatively high computational cost, especially on large problem sizes and data with a high proportion of outliers. In this paper, we propose a novel outlier removal technique for rotation search. Our method guarantees that any correspondence it discards as an outlier does not exist in the inlier set of the globally optimal rotation for the original data. Based on simple geometric operations, our algorithm is deterministic and fast. Used as a preprocessor to prune a large portion of the outliers from the input data, our method enables substantial speed-up of rotation search algorithms without compromising global optimality. We demonstrate the efficacy of our method in various synthetic and real data experiments.

The group membership prediction (GMP) problem involves predicting whether or not a collection of instances share a certain semantic property. For instance, in kinship verification given a collection of images, the goal is to predict whether or not they share a familial relationship. In this context we propose a novel probability model and introduce latent view-specific and view-shared random variables to jointly account for the view-specific appearance and cross-view similarities among data instances. Our model posits that data from each view is independent conditioned on the shared variables. This postulate leads to a parametric probability model that decomposes group membership likelihood into a tensor product of data-independent parameters and data-dependent factors. We propose learning the data-independent parameters in a discriminative way with bilinear classifiers, and test our prediction algorithm on challenging visual recognition tasks such as multi-camera person re-identification and kinship verification. On most benchmark datasets, our method can significantly outperform the current state-of-the-art.

We present a novel approach to 2D-3D registration from points or lines without correspondences. While there exist established solutions in the case where correspondences are known, there are many situations where it is not possible to reliably extract such correspondences across modalities, thus requiring the use of a correspondence-free registration algorithm. Existing correspondence-free methods rely on local search strategies and consequently have no guarantee of finding the optimal solution. In contrast, we present the first globally optimal approach to 2D-3D registration without correspondences, achieved by a Branch-and-Bound algorithm. Furthermore, a deterministic annealing procedure is proposed to speed up the nested branch-and-bound algorithm used. The theoretical and practical advantages this brings are demonstrated on a range of synthetic and real data where it is observed that the proposed approach is significantly more robust to high proportions of outliers compared to existing approaches.

Global structure-from-motion (SfM) methods solve all cameras simultaneously from all available relative motions. It has better potential in both reconstruction accuracy and computation efficiency than incremental methods. However, global SfM is challenging, mainly because of two reasons. Firstly, translation averaging is difficult, since an essential matrix only tells the direction of relative translation. Secondly, it is also hard to filter out bad essential matrices due to feature matching failures. We propose to compute a sparse depth image at each camera to solve both problems. Depth images help to upgrade an essential matrix to a similarity transformation, which can determine the scale of relative translation. Thus, camera registration is formulated as a well-posed similarity averaging problem. Depth images also make the filtering of essential matrices simple and effective. In this way, translation averaging can be solved robustly in two convex L1 optimization problems, which reach the global optimum rapidly. We demonstrate this method in various examples including sequential data, Internet data, and ambiguous data with repetitive scene structures.

Many recent efforts have been devoted to designing sophisticated deep learning structures, obtaining revolutionary results on benchmark datasets. The success of these deep learning methods mostly relies on an enormous volume of labeled training samples to learn a huge number of parameters in a network; therefore, understanding the generalization ability of a learned deep network cannot be overlooked, especially when restricted to a small training set, which is the case for many applications. In this paper, we propose a novel deep learning objective formulation that unifies both the classification and metric learning criteria. We then introduce a geometry-aware deep transform to enable a non-linear discriminative and robust feature transform, which shows competitive performance on small training sets for both synthetic and real-world data. We further support the proposed framework with a formal (K,epsilon)-robustness analysis.

In this work, we propose a generic scheme to promote any diffusion-based salient object detection algorithm by original ways to re-synthesize the diffusion matrix and construct the seed vector. We first make a novel analysis of the working mechanism of the diffusion matrix, which reveals the close relationship between saliency diffusion and spectral clustering. Following this analysis, we propose to re-synthesize the diffusion matrix from the most discriminative eigenvectors after adaptive re-weighting. Further, we propose to generate the seed vector based on the readily available diffusion maps, avoiding extra computation for color-based seed search. As a particular instance, we use inverse normalized Laplacian matrix as the original diffusion matrix and promote the corresponding salient object detection algorithm, which leads to superior performance as experimentally demonstrated.

We all have experienced forgetting habitual actions among our daily activities. For example, we probably have forgotten to turn the lights off before leaving a room or turn the stove off after cooking. In this paper, we propose a solution to the problem of issuing notifications on actions that may be missed. This involves learning about interdependencies between actions and being able to predict an ongoing action while segmenting the input video stream. In order to show a proof of concept, we collected a new egocentric dataset, in which people wear a camera while making lattes. We show promising results on the extremely challenging task of issuing correct and timely reminders. We also show that our model reliably segments the actions, while predicting the ongoing one when only a few frames from the beginning of the action are observed. The overall prediction accuracy is 46.2% when only 10 frames of an action are seen (2/3 of a sec). Moreover, the overall recognition and segmentation accuracy is shown to be 72.7% when the whole activity sequence is observed. Finally, the online prediction and segmentation accuracy is 68.3% when the prediction is made at every time step.

This paper introduces a general approach to dynamic scene reconstruction from multiple moving cameras without prior knowledge or limiting constraints on the scene structure, appearance, or illumination. Existing techniques or dynamic scene reconstruction from multiple wide-baseline camera views primarily focus on accurate reconstruction in controlled environments, where the cameras are fixed and calibrated and background is known. These approaches are not robust for general dynamic scenes captured with sparse moving cameras. Previous approaches for outdoor dynamic scene reconstruction assume prior knowledge of the static background appearance and structure. The primary contributions of this paper are twofold: an automatic method for initial coarse dynamic scene segmentation and reconstruction without prior knowledge of background appearance or structure; and a general robust approach for joint segmentation refinement and dense reconstruction of dynamic scenes from multiple wide-baseline static or moving cameras. Evaluation is performed on a variety of indoor and outdoor scenes with cluttered backgrounds and multiple dynamic non-rigid objects such as people. Comparison with state-of-the-art approaches demonstrates improved accuracy in both multiple view segmentation and dense reconstruction. The proposed approach also eliminates the requirement for prior knowledge of scene structure and appearance.

We present a novel approach to video segmentation using multiple object proposals. The problem is formulated as a minimization of a novel energy function defined over a fully connected graph of object proposals. Our model combines appearance with long-range point tracks, which is key to ensure robustness with respect to fast motion and occlusions over longer video sequences. As opposed to previous approaches based on object proposals, we do not seek the best per-frame object hypotheses to perform the segmentation. Instead, we combine multiple, potentially imperfect proposals to improve overall segmentation accuracy and ensure robustness to outliers. Overall, the basic algorithm consists of three steps. First, we generate a very large number of object proposals for each video frame using existing techniques. Next, we perform an SVM-based pruning step to retain only high quality proposals with sufficiently discriminative power. Finally, we determine the fore- and background classification by solving for the maximum a posteriori of a fully connected conditional random field, defined using our novel energy function. Experimental results on a well established dataset demonstrate that our method compares favorably to several recent state-of-the-art approaches.

This paper presents a linear time fully connected guided filter by introducing the minimum spanning tree (MST) to the guided filter (GF). Since the intensity based filtering kernel of GF is apt to overly smooth edges and the fixed-shape local box support region adopted by GF is not geometric-adaptive, our filter introduces an extra spatial term, the tree similarity, to the filtering kernel of GF and substitutes the box window with the implicit support region by establishing all-pairs-connections among pixels in the image and assigning the spatial-intensity-aware similarity to these connections. The adaptive implicit support region composed by the pixels with large kernel weights in the entire image domain has a big advantage over the predefined local box window in presenting the structure of an image for the reason that: 1, MST can efficiently present the structure of an image; 2, the kernel weight of our filter considers the tree distance defined on the MST. Due to these reasons, our filter achieves better edge-preserving results. We demonstrate the strength of the proposed filter in several applications. Experimental results show that our method produces better results than state-of-the-art methods.

In this paper, we propose a novel deep convolutional network (DCN) that achieves outstanding performance on FDDB, PASCAL Face, and AFW. Specifically, our method achieves a high recall rate of 90.99% on the challenging FDDB benchmark, outperforming the state-of-the-art method by a large margin of 2.91%. Importantly, we consider finding faces from a new perspective through scoring facial parts responses by their spatial structure and arrangement. The scoring mechanism is carefully formulated considering challenging cases where faces are only partially visible. This consideration allows our network to detect faces under severe occlusion and unconstrained pose variation, which are the main difficulty and bottleneck of most existing face detection approaches. We show that despite the use of DCN, our network can achieve practical runtime speed.

Limited annotated training data is a challenging problem in Action Unit recognition. In this paper, we investigate how the use of large databases labelled according to the 6 universal facial expressions can increase the generalization ability of Action Unit classifiers. For this purpose, we propose a novel learning framework: Hidden-Task Learning. HTL aims to learn a set of Hidden-Tasks (Action Units) for which samples are not available but, in contrast, training data is easier to obtain from a set of related Visible-Tasks (Facial Expressions). To that end, HTL is able to exploit prior knowledge about the relation between Hidden and Visible-Tasks. In our case, we base this prior knowledge on empirical psychological studies providing statistical correlations between Action Units and universal facial expressions. Additionally, we extend HTL to Semi-Hidden Task Learning (SHTL) assuming that Action Unit training samples are also provided. Performing exhaustive experiments over four different datasets, we show that HTL and SHTL improve the generalization ability of AU classifiers by training them with additional facial expression data. Additionally, we show that SHTL achieves competitive performance compared with state-of-the-art Transductive Learning approaches which face the problem of limited training data by using unlabelled test samples during training.

Extracting environment mattes using existing approaches often requires either thousands of captured images or a long processing time, or both. In this paper, we propose a novel approach to capturing and extracting the matte of a real scene effectively and efficiently. Grown out of the traditional frequency-based signal analysis, our approach can accurately locate contributing sources. By exploiting the recently developed compressive sensing theory, we simplify the data acquisition process of frequency-based environment matting. Incorporating phase information in a frequency signal into data acquisition further accelerates the matte extraction procedure. Compared with the state-of-the-art method, our approach achieves superior performance on both synthetic and real data, while consuming only a fraction of the processing time.