Silvia L. Pintea

Xin Liu, Fatemeh Karimi Nejadasl, Jan C. van Gemert et al.

Objects in videos are typically characterized by continuous smooth motion. We exploit continuous smooth motion in three ways. 1) Improved accuracy by using object motion as an additional source of supervision, which we obtain by anticipating object locations from a static keyframe. 2) Improved efficiency by only doing the expensive feature computations on a small subset of all frames. Because neighboring video frames are often redundant, we only compute features for a single static keyframe and predict object locations in subsequent frames. 3) Reduced annotation cost, where we only annotate the keyframe and use smooth pseudo-motion between keyframes. We demonstrate computational efficiency, annotation efficiency, and improved mean average precision compared to the state-of-the-art on four datasets: ImageNet VID, EPIC KITCHENS-55, YouTube-BoundingBoxes, and Waymo Open dataset. Our source code is available at https://github.com/L-KID/Videoobject-detection-by-location-anticipation.

Yancong Lin, Ruben Wiersma, Silvia L. Pintea et al.

Deep learning has improved vanishing point detection in images. Yet, deep networks require expensive annotated datasets trained on costly hardware and do not generalize to even slightly different domains, and minor problem variants. Here, we address these issues by injecting deep vanishing point detection networks with prior knowledge. This prior knowledge no longer needs to be learned from data, saving valuable annotation efforts and compute, unlocking realistic few-sample scenarios, and reducing the impact of domain changes. Moreover, the interpretability of the priors allows to adapt deep networks to minor problem variations such as switching between Manhattan and non-Manhattan worlds. We seamlessly incorporate two geometric priors: (i) Hough Transform -- mapping image pixels to straight lines, and (ii) Gaussian sphere -- mapping lines to great circles whose intersections denote vanishing points. Experimentally, we ablate our choices and show comparable accuracy to existing models in the large-data setting. We validate our model's improved data efficiency, robustness to domain changes, adaptability to non-Manhattan settings.

Silvia L. Pintea, Yancong Lin, Jouke Dijkstra et al.

A number of computer vision deep regression approaches report improved results when adding a classification loss to the regression loss. Here, we explore why this is useful in practice and when it is beneficial. To do so, we start from precisely controlled dataset variations and data samplings and find that the effect of adding a classification loss is the most pronounced for regression with imbalanced data. We explain these empirical findings by formalizing the relation between the balanced and imbalanced regression losses. Finally, we show that our findings hold on two real imbalanced image datasets for depth estimation (NYUD2-DIR), and age estimation (IMDB-WIKI-DIR), and on the problem of imbalanced video progress prediction (Breakfast). Our main takeaway is: for a regression task, if the data sampling is imbalanced, then add a classification loss.

Frans de Boer, Jan C. van Gemert, Jouke Dijkstra et al.

Activity progress prediction aims to estimate what percentage of an activity has been completed. Currently this is done with machine learning approaches, trained and evaluated on complicated and realistic video datasets. The videos in these datasets vary drastically in length and appearance. And some of the activities have unanticipated developments, making activity progression difficult to estimate. In this work, we examine the results obtained by existing progress prediction methods on these datasets. We find that current progress prediction methods seem not to extract useful visual information for the progress prediction task. Therefore, these methods fail to exceed simple frame-counting baselines. We design a precisely controlled dataset for activity progress prediction and on this synthetic dataset we show that the considered methods can make use of the visual information, when this directly relates to the progress prediction. We conclude that the progress prediction task is ill-posed on the currently used real-world datasets. Moreover, to fairly measure activity progression we advise to consider a, simple but effective, frame-counting baseline.

Silvia L. Pintea, Jouke Dijkstra

This work focuses on per-video unsupervised action segmentation, which is of interest to applications where storing large datasets is either not possible, or nor permitted. We propose to segment videos by learning in deep kernel space, to approximate the underlying frame distribution, as closely as possible. To define this closeness metric between the original video distribution and its approximation, we rely on maximum mean discrepancy (MMD) which is a geometry-preserving metric in distribution space, and thus gives more reliable estimates. Moreover, unlike the commonly used optimal transport metric, MMD is both easier to optimize, and faster. We choose to use neural tangent kernels (NTKs) to define the kernel space where MMD operates, because of their improved descriptive power as opposed to fixed kernels. And, also, because NTKs sidestep the trivial solution, when jointly learning the inputs (video approximation) and the kernel function. Finally, we show competitive results when compared to state-of-the-art per-video methods, on six standard benchmarks. Additionally, our method has higher F1 scores than prior agglomerative work, when the number of segments is unknown.

Xiaotong Zhang, Alexander Broersen, Gonnie CM van Erp et al.

The preoperative planning of liver surgery relies on Couinaud segmentation from computed tomography (CT) images, to reduce the risk of bleeding and guide the resection procedure. Using 3D point-based representations, rather than voxelizing the CT volume, has the benefit of preserving the physical resolution of the CT. However, point-based representations need prior knowledge of the liver vessel structure, which is time consuming to acquire. Here, we propose a point-based method for Couinaud segmentation, without explicitly providing the prior liver vessel structure. To allow the model to learn this anatomical liver vessel structure, we add a graph reasoning module on top of the point features. This adds implicit anatomical information to the model, by learning affinities across point neighborhoods. Our method is competitive on the MSD and LiTS public datasets in Dice coefficient and average surface distance scores compared to four pioneering point-based methods. Our code is available at https://github.com/ZhangXiaotong015/GrPn.

Xiaotong Zhang, Alexander Broersen, Gonnie CM van Erp et al.

Improving connectivity and completeness are the most challenging aspects of liver vessel segmentation, especially for small vessels. These challenges require both learning the continuous vessel geometry, and focusing on small vessel detection. However, current methods do not explicitly address these two aspects and cannot generalize well when constrained by inconsistent annotations. Here, we take advantage of the generalization of the diffusion model and explicitly integrate connectivity and completeness in our diffusion-based segmentation model. Specifically, we use a graph-attention module that adds knowledge about vessel geometry, and thus adds continuity. Additionally, we perform the graph-attention at multiple-scales, thus focusing on small liver vessels. Our method outperforms eight state-of-the-art medical segmentation methods on two public datasets: 3D-ircadb-01 and LiVS. Our code is available at https://github.com/ZhangXiaotong015/GATSegDiff.

Nergis Tomen, Silvia L. Pintea, Jan C. van Gemert

CNNs and computational models of biological vision share some fundamental principles, which opened new avenues of research. However, fruitful cross-field research is hampered by conventional CNN architectures being based on spatially and depthwise discrete representations, which cannot accommodate certain aspects of biological complexity such as continuously varying receptive field sizes and dynamics of neuronal responses. Here we propose deep continuous networks (DCNs), which combine spatially continuous filters, with the continuous depth framework of neural ODEs. This allows us to learn the spatial support of the filters during training, as well as model the continuous evolution of feature maps, linking DCNs closely to biological models. We show that DCNs are versatile and highly applicable to standard image classification and reconstruction problems, where they improve parameter and data efficiency, and allow for meta-parametrization. We illustrate the biological plausibility of the scale distributions learned by DCNs and explore their performance in a neuroscientifically inspired pattern completion task. Finally, we investigate an efficient implementation of DCNs by changing input contrast.

Xiaotong Zhang, Alexander Broersen, Gonnie CM van Erp et al.

Liver-vessel segmentation is an essential task in the pre-operative planning of liver resection. State-of-the-art 2D or 3D convolution-based methods focusing on liver vessel segmentation on 2D CT cross-sectional views, which do not take into account the global liver-vessel topology. To maintain this global vessel topology, we rely on the underlying physics used in the CT reconstruction process, and apply this to liver-vessel segmentation. Concretely, we introduce the concept of top-k maximum intensity projections, which mimics the CT reconstruction by replacing the integral along each projection direction, with keeping the top-k maxima along each projection direction. We use these top-k maximum projections to condition a diffusion model and generate 3D liver-vessel trees. We evaluate our 3D liver-vessel segmentation on the 3D-ircadb-01 dataset, and achieve the highest Dice coefficient, intersection-over-union (IoU), and Sensitivity scores compared to prior work.

Yunqiang Li, Silvia L. Pintea, Jan C. van Gemert

Binary networks are extremely efficient as they use only two symbols to define the network: $\{+1,-1\}$. One can make the prior distribution of these symbols a design choice. The recent IR-Net of Qin et al. argues that imposing a Bernoulli distribution with equal priors (equal bit ratios) over the binary weights leads to maximum entropy and thus minimizes information loss. However, prior work cannot precisely control the binary weight distribution during training, and therefore cannot guarantee maximum entropy. Here, we show that quantizing using optimal transport can guarantee any bit ratio, including equal ratios. We investigate experimentally that equal bit ratios are indeed preferable and show that our method leads to optimization benefits. We show that our quantization method is effective when compared to state-of-the-art binarization methods, even when using binary weight pruning.

Nikhil Saldanha, Silvia L. Pintea, Jan C. van Gemert et al.

Frequency information lies at the base of discriminating between textures, and therefore between different objects. Classical CNN architectures limit the frequency learning through fixed filter sizes, and lack a way of explicitly controlling it. Here, we build on the structured receptive field filters with Gaussian derivative basis. Yet, rather than using predetermined derivative orders, which typically result in fixed frequency responses for the basis functions, we learn these. We show that by learning the order of the basis we can accurately learn the frequency of the filters, and hence adapt to the optimal frequencies for the underlying learning task. We investigate the well-founded mathematical formulation of fractional derivatives to adapt the filter frequencies during training. Our formulation leads to parameter savings and data efficiency when compared to the standard CNNs and the Gaussian derivative CNN filter networks that we build upon.

Silvia L. Pintea, Nergis Tomen, Stanley F. Goes et al.

Resolution in deep convolutional neural networks (CNNs) is typically bounded by the receptive field size through filter sizes, and subsampling layers or strided convolutions on feature maps. The optimal resolution may vary significantly depending on the dataset. Modern CNNs hard-code their resolution hyper-parameters in the network architecture which makes tuning such hyper-parameters cumbersome. We propose to do away with hard-coded resolution hyper-parameters and aim to learn the appropriate resolution from data. We use scale-space theory to obtain a self-similar parametrization of filters and make use of the N-Jet: a truncated Taylor series to approximate a filter by a learned combination of Gaussian derivative filters. The parameter sigma of the Gaussian basis controls both the amount of detail the filter encodes and the spatial extent of the filter. Since sigma is a continuous parameter, we can optimize it with respect to the loss. The proposed N-Jet layer achieves comparable performance when used in state-of-the art architectures, while learning the correct resolution in each layer automatically. We evaluate our N-Jet layer on both classification and segmentation, and we show that learning sigma is especially beneficial for inputs at multiple sizes.

Xin Liu, Silvia L. Pintea, Fatemeh Karimi Nejadasl et al.

Not all video frames are equally informative for recognizing an action. It is computationally infeasible to train deep networks on all video frames when actions develop over hundreds of frames. A common heuristic is uniformly sampling a small number of video frames and using these to recognize the action. Instead, here we propose full video action recognition and consider all video frames. To make this computational tractable, we first cluster all frame activations along the temporal dimension based on their similarity with respect to the classification task, and then temporally aggregate the frames in the clusters into a smaller number of representations. Our method is end-to-end trainable and computationally efficient as it relies on temporally localized clustering in combination with fast Hamming distances in feature space. We evaluate on UCF101, HMDB51, Breakfast, and Something-Something V1 and V2, where we compare favorably to existing heuristic frame sampling methods.

Yancong Lin, Silvia L. Pintea, Jan C. van Gemert

Classical work on line segment detection is knowledge-based; it uses carefully designed geometric priors using either image gradients, pixel groupings, or Hough transform variants. Instead, current deep learning methods do away with all prior knowledge and replace priors by training deep networks on large manually annotated datasets. Here, we reduce the dependency on labeled data by building on the classic knowledge-based priors while using deep networks to learn features. We add line priors through a trainable Hough transform block into a deep network. Hough transform provides the prior knowledge about global line parameterizations, while the convolutional layers can learn the local gradient-like line features. On the Wireframe (ShanghaiTech) and York Urban datasets we show that adding prior knowledge improves data efficiency as line priors no longer need to be learned from data. Keywords: Hough transform; global line prior, line segment detection.

Ioannis Lelekas, Nergis Tomen, Silvia L. Pintea et al.

Biological vision adopts a coarse-to-fine information processing pathway, from initial visual detection and binding of salient features of a visual scene, to the enhanced and preferential processing given relevant stimuli. On the contrary, CNNs employ a fine-to-coarse processing, moving from local, edge-detecting filters to more global ones extracting abstract representations of the input. In this paper we reverse the feature extraction part of standard bottom-up architectures and turn them upside-down: We propose top-down networks. Our proposed coarse-to-fine pathway, by blurring higher frequency information and restoring it only at later stages, offers a line of defence against adversarial attacks that introduce high frequency noise. Moreover, since we increase image resolution with depth, the high resolution of the feature map in the final convolutional layer contributes to the explainability of the network's decision making process. This favors object-driven decisions over context driven ones, and thus provides better localized class activation maps. This paper offers empirical evidence for the applicability of the top-down resolution processing to various existing architectures on multiple visual tasks.

Omar Hommos, Silvia L. Pintea, Pascal S. M. Mettes et al.

Currently, the most common motion representation for action recognition is optical flow. Optical flow is based on particle tracking which adheres to a Lagrangian perspective on dynamics. In contrast to the Lagrangian perspective, the Eulerian model of dynamics does not track, but describes local changes. For video, an Eulerian phase-based motion representation, using complex steerable filters, has been successfully employed recently for motion magnification and video frame interpolation. Inspired by these previous works, here, we proposes learning Eulerian motion representations in a deep architecture for action recognition. We learn filters in the complex domain in an end-to-end manner. We design these complex filters to resemble complex Gabor filters, typically employed for phase-information extraction. We propose a phase-information extraction module, based on these complex filters, that can be used in any network architecture for extracting Eulerian representations. We experimentally analyze the added value of Eulerian motion representations, as extracted by our proposed phase extraction module, and compare with existing motion representations based on optical flow, on the UCF101 dataset.

Silvia L. Pintea, Jian Zheng, Xilin Li et al.

We focus on the problem of estimating human hand-tremor frequency from input RGB video data. Estimating tremors from video is important for non-invasive monitoring, analyzing and diagnosing patients suffering from motor-disorders such as Parkinson's disease. We consider two approaches for hand-tremor frequency estimation: (a) a Lagrangian approach where we detect the hand at every frame in the video, and estimate the tremor frequency along the trajectory; and (b) an Eulerian approach where we first localize the hand, we subsequently remove the large motion along the movement trajectory of the hand, and we use the video information over time encoded as intensity values or phase information to estimate the tremor frequency. We estimate hand tremors on a new human tremor dataset, TIM-Tremor, containing static tasks as well as a multitude of more dynamic tasks, involving larger motion of the hands. The dataset has 55 tremor patient recordings together with: associated ground truth accelerometer data from the most affected hand, RGB video data, and aligned depth data.

Silvia L. Pintea, Yue Liu, Jan C. van Gemert

Knowledge distillation compacts deep networks by letting a small student network learn from a large teacher network. The accuracy of knowledge distillation recently benefited from adding residual layers. We propose to reduce the size of the student network even further by recasting multiple residual layers in the teacher network into a single recurrent student layer. We propose three variants of adding recurrent connections into the student network, and show experimentally on CIFAR-10, Scenes and MiniPlaces, that we can reduce the number of parameters at little loss in accuracy.

Silvia L. Pintea, Pascal S. Mettes, Jan C. van Gemert et al.

This method introduces an efficient manner of learning action categories without the need of feature estimation. The approach starts from low-level values, in a similar style to the successful CNN methods. However, rather than extracting general image features, we learn to predict specific video representations from raw video data. The benefit of such an approach is that at the same computational expense it can predict 2 D video representations as well as 3 D ones, based on motion. The proposed model relies on discriminative Waldboost, which we enhance to a multiclass formulation for the purpose of learning video representations. The suitability of the proposed approach as well as its time efficiency are tested on the UCF11 action recognition dataset.

Silvia L. Pintea, Jan C. van Gemert, Arnold W. M. Smeulders

This work incorporates the multi-modality of the data distribution into a Gaussian Process regression model. We approach the problem from a discriminative perspective by learning, jointly over the training data, the target space variance in the neighborhood of a certain sample through metric learning. We start by using data centers rather than all training samples. Subsequently, each center selects an individualized kernel metric. This enables each center to adjust the kernel space in its vicinity in correspondence with the topology of the targets --- a multi-modal approach. We additionally add descriptiveness by allowing each center to learn a precision matrix. We demonstrate empirically the reliability of the model.

Silvia L. Pintea, Jan C. van Gemert, Arnold W. M. Smeulders

This paper proposes motion prediction in single still images by learning it from a set of videos. The building assumption is that similar motion is characterized by similar appearance. The proposed method learns local motion patterns given a specific appearance and adds the predicted motion in a number of applications. This work (i) introduces a novel method to predict motion from appearance in a single static image, (ii) to that end, extends of the Structured Random Forest with regression derived from first principles, and (iii) shows the value of adding motion predictions in different tasks such as: weak frame-proposals containing unexpected events, action recognition, motion saliency. Illustrative results indicate that motion prediction is not only feasible, but also provides valuable information for a number of applications.

Yichao Zhang, Silvia L. Pintea, Jan C. van Gemert

The ability to amplify or reduce subtle image changes over time is useful in contexts such as video editing, medical video analysis, product quality control and sports. In these contexts there is often large motion present which severely distorts current video amplification methods that magnify change linearly. In this work we propose a method to cope with large motions while still magnifying small changes. We make the following two observations: i) large motions are linear on the temporal scale of the small changes; ii) small changes deviate from this linearity. We ignore linear motion and propose to magnify acceleration. Our method is pure Eulerian and does not require any optical flow, temporal alignment or region annotations. We link temporal second-order derivative filtering to spatial acceleration magnification. We apply our method to moving objects where we show motion magnification and color magnification. We provide quantitative as well as qualitative evidence for our method while comparing to the state-of-the-art.