We present a system to create Mobile Realistic Fullbody (MoRF) avatars. MoRF avatars are rendered in real-time on mobile devices, learned from monocular videos, and have high realism. We use SMPL-X as a proxy geometry and render it with DNR (neural texture and image-2-image network). We improve on prior work, by overfitting perframe warping fields in the neural texture space, allowing to better align the training signal between different frames. We also refine SMPL-X mesh fitting procedure to improve the overall avatar quality. In the comparisons to other monocular video-based avatar systems, MoRF avatars achieve higher image sharpness and temporal consistency. Participants of our user study also preferred avatars generated by MoRF.

Generating realistic human 3D reconstructions using image or video data is essential for various communication and entertainment applications. While existing methods achieved impressive results for body and facial regions, realistic hair modeling still remains challenging due to its high mechanical complexity. This work proposes an approach capable of accurate hair geometry reconstruction at a strand level from a monocular video or multi-view images captured in uncontrolled lighting conditions. Our method has two stages, with the first stage performing joint reconstruction of coarse hair and bust shapes and hair orientation using implicit volumetric representations. The second stage then estimates a strand-level hair reconstruction by reconciling in a single optimization process the coarse volumetric constraints with hair strand and hairstyle priors learned from the synthetic data. To further increase the reconstruction fidelity, we incorporate image-based losses into the fitting process using a new differentiable renderer. The combined system, named Neural Haircut, achieves high realism and personalization of the reconstructed hairstyles. For video results, please refer to our project page†.

We present DINAR, an approach for creating realistic rigged fullbody avatars from single RGB images. Similarly to previous works, our method uses neural textures combined with the SMPL-X body model to achieve photorealistic quality of avatars while keeping them easy to animate and fast to infer. To restore the texture, we use a latent diffusion model and show how such model can be trained in the neural texture space. The use of the diffusion model allows us to realistically reconstruct large unseen regions such as the back of a person given the frontal view. The models in our pipeline are trained using 2D images and videos only. In the experiments, our approach achieves state-of-the-art rendering quality and good generalization to new poses and viewpoints. In particular, the approach improves state-of-the-art on the SnapshotPeople public benchmark.

Representing scenes with multiple semitransparent colored layers has been a popular and successful choice for real-time novel view synthesis. Existing approaches infer colors and transparency values over regularly spaced layers of planar or spherical shape. In this work, we introduce a new view synthesis approach based on multiple semitransparent layers with scene-adapted geometry. Our approach infers such representations from stereo pairs in two stages. The first stage produces the geometry of a small number of data-adaptive layers from a given pair of views. The second stage infers the color and transparency values for these layers, producing the final representation for novel view synthesis. Importantly, both stages are connected through a differentiable renderer and are trained end-to-end. In the experiments, we demonstrate the advantage of the proposed approach over the use of regularly spaced layers without adaptation to scene geometry. Despite being orders of magnitude faster during rendering, our approach also outperforms the recently proposed IBRNet system based on implicit geometry representation.

We present a new system$(NPBG++)$for the novel view synthesis (NVS) task that achieves high rendering realism with low scene fitting time. Our method efficiently lever-ages the multiview observations and the point cloud of a static scene to predict a neural descriptor for each point, improving upon the pipeline of Neural Point-Based Graph-ics [1] in several important ways. By predicting the descrip-tors with a single pass through the source images, we lift the requirement of per-scene optimization while also making the neural descriptors view-dependent and more suit-able for scenes with strong non-Lambertian effects. In our comparisons, the proposed system outperforms previous NVS approaches in terms of fitting and rendering runtimes while producing images of similar quality. Project page: https://rakhimovv.github.io/npbgpp/.

In this work, we advance the neural head avatar technology to the megapixel resolution while focusing on the particularly challenging task of cross-driving synthesis, i.e., when the appearance of the driving image is substantially different from the animated source image. We propose a set of new neural architectures and training methods that can leverage both medium-resolution video data and high-resolution image data to achieve the desired levels of rendered image quality and generalization to novel views and motion. We demonstrate that suggested architectures and methods produce convincing high-resolution neural avatars, outperforming the competitors in the cross-driving scenario. Lastly, we show how a trained high-resolution neural avatar model can be distilled into a lightweight student model which runs in real-time and locks the identities of neural avatars to several dozens of pre-defined source images. Real-time operation and identity lock are essential for many practical applications head avatar systems.

Modern image inpainting systems, despite the significant progress, often struggle with large missing areas, complex geometric structures, and high-resolution images. We find that one of the main reasons for that is the lack of an effective receptive field in both the inpainting network and the loss function. To alleviate this issue, we propose a new method called large mask inpainting (LaMa). LaMa is based on i) a new inpainting network architecture that uses fast Fourier convolutions (FFCs), which have the image-wide receptive field; ii) a high receptive field perceptual loss; iii) large training masks, which unlocks the potential of the first two components. Our inpainting network improves the state-of-the-art across a range of datasets and achieves excellent performance even in challenging scenarios, e.g. completion of periodic structures. Our model generalizes surprisingly well to resolutions that are higher than those seen at train time, and achieves this at lower parameter&time costs than the competitive baselines. The code is available at https://github.com/saic-mdal/lama.

Existing image generator networks rely heavily on spatial convolutions and, optionally, self-attention blocks in order to gradually synthesize images in a coarse-to-fine manner. Here, we present a new architecture for image generators, where the color value at each pixel is computed independently given the value of a random latent vector and the coordinate of that pixel. No spatial convolutions or similar operations that propagate information across pixels are involved during the synthesis. We analyze the modeling capabilities of such generators when trained in an adversarial fashion, and observe the new generators to achieve similar generation quality to state-of-the-art convolutional generators. We also investigate several interesting properties unique to the new architecture.

We propose a new type of full-body human avatars, which combines parametric mesh-based body model with a neural texture. We show that with the help of neural textures, such avatars can successfully model clothing and hair, which usually poses a problem for mesh-based approaches. We also show how these avatars can be created from multiple frames of a video using backpropagation. We then propose a generative model for such avatars that can be trained from datasets of images and videos of people. The generative model allows us to sample random avatars as well as to create dressed avatars of people from one or few images. The code for the project is available at saic-violet.github.io/style-people.

We present a new versatile building block for deep point cloud processing architectures that is equally suited for diverse tasks. This building block combines the ideas of spatial transformers and multi-view convolutional networks with the efficiency of standard convolutional layers in two and three-dimensional dense grids. The new block operates via multiple parallel heads, whereas each head differentiably rasterizes feature representations of individual points into a low-dimensional space, and then uses dense convolution to propagate information across points. The results of the processing of individual heads are then combined together resulting in the update of point features. Using the new block, we build architectures for both discriminative (point cloud segmentation, point cloud classification) and generative (point cloud inpainting and image-based point cloud reconstruction) tasks. The resulting architectures achieve state-of-the-art performance for these tasks, demonstrating the versatility of the new block for point cloud processing.

We propose a new approach to human clothing modeling based on point clouds. Within this approach, we learn a deep model that can predict point clouds of various outfits, for various human poses, and for various human body shapes. Notably, outfits of various types and topologies can be handled by the same model. Using the learned model, we can infer the geometry of new outfits from as little as a single image, and perform outfit retargeting to new bodies in new poses. We complement our geometric model with appearance modeling that uses the point cloud geometry as a geometric scaffolding and employs neural point-based graphics to capture outfit appearance from videos and to re-render the captured outfits. We validate both geometric modeling and appearance modeling aspects of the proposed approach against recently proposed methods and establish the viability of point-based clothing modeling.

We present a system for real-time RGBD-based estimation of 3D human pose. We use parametric 3D deformable human mesh model (SMPL-X) as a representation and focus on the real-time estimation of parameters for the body pose, hands pose and facial expression from Kinect Azure RGB-D camera. We train estimators of body pose and facial expression parameters. Both estimators use previously published landmark extractors as input and custom annotated datasets for supervision, while hand pose is estimated directly by a previously published method. We combine the predictions of those estimators into a temporally-smooth human pose. We train the facial expression extractor on a large talking face dataset, which we annotate with facial expression parameters. For the body pose we collect and annotate a dataset of 56 people captured from a rig of 5 Kinect Azure RGB-D cameras and use it together with a large motion capture AMASS dataset. Our RGB-D body pose model outperforms the state-of-the-art RGB-only methods and works on the same level of accuracy compared to a slower RGB-D optimization-based solution. The combined system runs at 25 FPS on a server with a single GPU. The code will be available at saic-violet.github.io/rgbd-kinect-pose.

We propose and evaluate a neural point-based graphics method that can model semi-transparent scene parts. Similarly to its predecessor pipeline, ours uses point clouds to model proxy geometry, and augments each point with a neural descriptor. Additionally, a learnable transparency value is introduced in our approach for each point. Our neural rendering procedure consists of two steps. Firstly, the point cloud is rasterized using ray marching into a multi-channel image. This is followed by the neural rendering step that “translates” the rasterized image into an RGB output using a learnable convolutional network. New scenes can be modeled using gradient-based optimization of neural descriptors and of the rendering network.We show that novel views of semi-transparent point cloud scenes can be generated after training with our approach. Our experiments demonstrate the benefit of introducing semi-transparency into the neural point-based modeling for a range of scenes with semi-transparent parts. The project materials and the code are available at http://saic-violet.github.io/transpr.

Modeling daytime changes in high resolution photographs, e.g., re-rendering the same scene under different illuminations typical for day, night, or dawn, is a challenging image manipulation task. We present the high-resolution daytime translation (HiDT) model for this task. HiDT combines a generative image-to-image model and a new upsampling scheme that allows to apply image translation at high resolution. The model demonstrates competitive results in terms of both commonly used GAN metrics and human evaluation. Importantly, this good performance comes as a result of training on a dataset of still landscape images with no daytime labels available. Our results are available at https://saic-mdal.github.io/HiDT/.

We propose a neural head reenactment system, which is driven by a latent pose representation and is capable of predicting the foreground segmentation alongside the RGB image. The latent pose representation is learned as a part of the entire reenactment system, and the learning process is based solely on image reconstruction losses. We show that despite its simplicity, with a large and diverse enough training dataset, such learning successfully decomposes pose from identity. The resulting system can then reproduce mimics of the driving person and, furthermore, can perform cross-person reenactment. Additionally, we show that the learned descriptors are useful for other pose-related tasks, such as keypoint prediction and pose-based retrieval.

Computer vision tasks such as image classification, image retrieval, and few-shot learning are currently dominated by Euclidean and spherical embeddings so that the final decisions about class belongings or the degree of similarity are made using linear hyperplanes, Euclidean distances, or spherical geodesic distances (cosine similarity). In this work, we demonstrate that in many practical scenarios, hyperbolic embeddings provide a better alternative.

We present a new instance segmentation approach tailored to biological images, where instances may correspond to individual cells, organisms or plant parts. Unlike instance segmentation for user photographs or road scenes, in biological data object instances may be particularly densely packed, the appearance variation may be particularly low, the processing power may be restricted, while, on the other hand, the variability of sizes of individual instances may be limited. The proposed approach successfully addresses these peculiarities. Our approach describes each object instance using an expectation of a limited number of sine waves with frequencies and phases adjusted to particular object sizes and densities. At train time, a fully-convolutional network is learned to predict the object embeddings at each pixel using a simple pixelwise regression loss, while at test time the instances are recovered using clustering in the embedding space. In the experiments, we show that our approach outperforms previous embedding-based instance segmentation approaches on a number of biological datasets, achieving state-of-the-art on a popular CVPPP benchmark. This excellent performance is combined with computational efficiency that is needed for deployment to domain specialists. The source code of the approach is available at https://github.com/kulikovv/harmonic .

We present a new deep learning approach to pose-guided resynthesis of human photographs. At the heart of the new approach is the estimation of the complete body surface texture based on a single photograph. Since the input photograph always observes only a part of the surface, we suggest a new inpainting method that completes the texture of the human body. Rather than working directly with colors of texture elements, the inpainting network estimates an appropriate source location in the input image for each element of the body surface. This correspondence field between the input image and the texture is then further warped into the target image coordinate frame based on the desired pose, effectively establishing the correspondence between the source and the target view even when the pose change is drastic. The final convolutional network then uses the established correspondence and all other available information to synthesize the output image. A fully-convolutional architecture with deformable skip connections guided by the estimated correspondence field is used. We show state-of-the-art result for pose-guided image synthesis. Additionally, we demonstrate the performance of our system for garment transfer and pose-guided face resynthesis.

We present a system for learning full body neural avatars, i.e. deep networks that produce full body renderings of a person for varying body pose and varying camera pose. Our system takes the middle path between the classical graphics pipeline and the recent deep learning approaches that generate images of humans using image-to-image translation. In particular, our system estimates an explicit two-dimensional texture map of the model surface. At the same time, it abstains from explicit shape modeling in 3D. Instead, at test time, the system uses a fully-convolutional network to directly map the configuration of body feature points w.r.t. the camera to the 2D texture coordinates of individual pixels in the image frame. We show that such system is capable of learning to generate realistic renderings while being trained on videos annotated with 3D poses and foreground masks. We also demonstrate that maintaining an explicit texture representation helps our system to achieve better generalization compared to systems that use direct image-to-image translation.

We present two novel solutions for multi-view 3D human pose estimation based on new learnable triangulation methods that combine 3D information from multiple 2D views. The first (baseline) solution is a basic differentiable algebraic triangulation with an addition of confidence weights estimated from the input images. The second solution is based on a novel method of volumetric aggregation from intermediate 2D backbone feature maps. The aggregated volume is then refined via 3D convolutions that produce final 3D joint heatmaps and allow implicit modelling a human pose prior. Crucially, both approaches are end-to-end differentiable, which allows us to directly optimize the target metric. We demonstrate transferability of the solutions across datasets and considerably improve the multiview state of the art on the Human3.6M dataset. Video demonstration, annotations and additional materials will be posted on our project page.

Several recent works have shown how highly realistic human head images can be obtained by training convolutional neural networks to generate them. In order to create a personalized talking head model, these works require training on a large dataset of images of a single person. However, in many practical scenarios, such personalized talking head models need to be learned from a few image views of a person, potentially even a single image. Here, we present a system with such few-shot capability. It performs lengthy meta-learning on a large dataset of videos, and after that is able to frame few- and one-shot learning of neural talking head models of previously unseen people as adversarial training problems with high capacity generators and discriminators. Crucially, the system is able to initialize the parameters of both the generator and the discriminator in a person-specific way, so that training can be based on just a few images and done quickly, despite the need to tune tens of millions of parameters. We show that such an approach is able to learn highly realistic and personalized talking head models of new people and even portrait paintings.

Neuronal synapses transmit electrochemical signals between cells through the coordinated action of presynaptic vesicles, ion channels, scaffolding and adapter proteins, and membrane receptors. In situ structural characterization of numerous synaptic proteins simultaneously through multiplexed imaging facilitates a bottom-up approach to synapse classification and phenotypic description. Objective automation of efficient and reliable synapse detection within these datasets is essential for the high-throughput investigation of synaptic features. Convolutional neural networks can solve this generalized problem of synapse detection, however, these architectures require large numbers of training examples to optimize their thousands of parameters. We propose DoGNet, a neural network architecture that closes the gap between classical computer vision blob detectors, such as Difference of Gaussians (DoG) filters, and modern convolutional networks. DoGNet is optimized to analyze highly multiplexed microscopy data. Its small number of training parameters allows DoGNet to be trained with few examples, which facilitates its application to new datasets without overfitting. We evaluate the method on multiplexed fluorescence imaging data from both primary mouse neuronal cultures and mouse cortex tissue slices. We show that DoGNet outperforms convolutional networks with a low-to-moderate number of training examples, and DoGNet is efficiently transferred between datasets collected from separate research groups. DoGNet synapse localizations can then be used to guide the segmentation of individual synaptic protein locations and spatial extents, revealing their spatial organization and relative abundances within individual synapses. The source code is publicly available: https://github.com/kulikovv/dognet.

We present a new autoencoder-type architecture that is trainable in an unsupervised mode, sustains both generation and inference, and has the quality of conditional and unconditional samples boosted by adversarial learning. Unlike previous hybrids of autoencoders and adversarial networks, the adversarial game in our approach is set up directly between the encoder and the generator, and no external mappings are trained in the process of learning.The game objective compares the divergences of each of the real and the generated data distributions with the prior distribution in the latent space. We show that direct generator-vs-encoder game leads to a tight coupling of the two components, resulting in samples and reconstructions of a comparable quality to some recently-proposed more complex architectures.

Deep convolutional networks have become a popular tool for image generation and restoration. Generally, their excellent performance is imputed to their ability to learn realistic image priors from a large number of example images. In this paper, we show that, on the contrary, the structure of a generator network is sufficient to capture a great deal of low-level image statistics prior to any learning. In order to do so, we show that a randomly-initialized neural network can be used as a handcrafted prior with excellent results in standard inverse problems such as denoising, superresolution, and inpainting. Furthermore, the same prior can be used to invert deep neural representations to diagnose them, and to restore images based on flash-no flash input pairs. Apart from its diverse applications, our approach highlights the inductive bias captured by standard generator network architectures. It also bridges the gap between two very popular families of image restoration methods: learning-based methods using deep convolutional networks and learning-free methods based on handcrafted image priors such as self-similarity.

We present a new approach to monocular learning-based gaze redirection problem in images that is able to train on raw sequences of eye images with unknown gaze directions and a small amount of eye images, where the gaze direction is known. The proposed approach is based on a pair of deep networks, where the first encoder-like network maps eye images to a latent space, while the second network maps pairs of latent representations to warping fields implementing the transformation between the pair of the original images. In the proposed system, both networks are trained in an unsupervised manner, while the gaze-annotated images are only used to estimate displacements in the latent space that are characteristic to certain gaze redirections. Quantitative and qualitative evaluation suggests that such characteristic displacement vectors in the learned latent space can be learned from few examples and are transferable across different people and different imaging conditions.

Box filters computed using integral images have been part of the computer vision toolset for a long time. Here, we show that a convolutional layer that computes box filter responses in a sliding manner can be used within deep architectures, whereas the dimensions and the offsets of the sliding boxes in such a layer can be learned as part of an end-to-end loss minimization. Crucially, the training process can make the size of the boxes in such a layer arbitrarily large without incurring extra computational cost and without the need to increase the number of learnable parameters. Due to its ability to integrate information over large boxes, the new layer facilitates long-range propagation of information and leads to the efficient increase of the receptive fields of downstream units in the network. By incorporating the new layer into existing architectures for semantic segmentation, we are able to achieve both the increase in segmentation accuracy as well as the decrease in the computational cost and the number of learnable parameters.

We present CarpetVR - a new system for marker-based positional tracking suitable for mobile VR (Figure 1). The system utilizes all sensors present on a modern smartphone (a camera, a gyroscope, and an accelerometer) and does not require any additional sensors. CarpetVR uses a single floor marker that we call the magic carpet. CarpetVR augments a standard mobile VR setup with a slanted mirror that can be attached either to the smartphone or to the head mount in front of the smartphone camera. As the person walks over the marker, the smartphone camera is able to see the marker thanks to the reflection in the mirror. Our tracking engine then uses a computer vision module to detect the marker and to estimate the smartphone position with respect to the marker at 40 frames per second. This estimate is integrated with high framerate signals from the gyroscope and the accelerometer. The resulting estimates of the position and the orientation are then used to render the virtual world. Our sensor fusion algorithm ensures minimal-latency tracking with very little jitter.

We propose a general approach to the gaze redirection problem in images that utilizes machine learning. The idea is to learn to re-synthesize images by training on pairs of images with known disparities between gaze directions. We show that such learning-based re-synthesis can achieve convincing gaze redirection based on monocular input, and that the learned systems generalize well to people and imaging conditions unseen during training. We describe and compare three instantiations of our idea. The first system is based on efficient decision forest predictors and redirects the gaze by a fixed angle in real-time (on a single CPU), being particularly suitable for the videoconferencing gaze correction. The second system is based on a deep architecture and allows gaze redirection by a range of angles. The second system achieves higher photorealism, while being several times slower. The third system is based on real-time decision forests at test time, while using the supervision from a "teacher" deep network during training. The third system approaches the quality of a teacher network in our experiments, and thus provides a highly realistic real-time monocular solution to the gaze correction problem. We present in-depth assessment and comparisons of the proposed systems based on quantitative measurements and a user study.

In this work we propose a new architecture for person re-identification. As the task of re-identification is inherently associated with embedding learning and non-rigid appearance description, our architecture is based on the deep bilinear convolutional network (Bilinear-CNN) that has been proposed recently for fine-grained classification of highly non-rigid objects. While the last stages of the original Bilinear-CNN architecture completely removes the geometric information from consideration by performing orderless pooling, we observe that a better embedding can be learned by performing bilinear pooling in a more local way, where each pooling is confined to a predefined region. Our architecture thus represents a compromise between traditional convolutional networks and bilinear CNNs and strikes a balance between rigid matching and completely ignoring spatial information. We perform the experimental validation of the new architecture on the three popular benchmark datasets (Market-1501, CUHK01, CUHK03), comparing it to baselines that include Bilinear-CNN as well as prior art. The new architecture outperforms the baseline on all three datasets, while performing better than state-of-the-art on two out of three. The code and the pretrained models of the approach will be made available at the time of publication.

In this work, we introduce a new kind of spatial partition trees for efficient nearest-neighbor search. Our approach first identifies a set of useful data splitting directions, and then learns a codebook that can be used to encode such directions. We use the product-quantization idea in order to make the effective codebook large, the evaluation of scalar products between the query and the encoded splitting direction very fast, and the encoding itself compact. As a result, the proposed data srtucture (Product Split tree) achieves compact clustering of data points, while keeping the traversal very efficient. In the nearest-neighbor search experiments on high-dimensional data, product split trees achieved state-of-the-art performance, demonstrating better speed-accuracy tradeoff than other spatial partition trees.

This work is motivated by the mostly unsolved task of parsing biological images with multiple overlapping articulated model organisms (such as worms or larvae). We present a general approach that separates the two main challenges associated with such data, individual object shape estimation and object groups disentangling. At the core of the approach is a deep feed-forward singling-out network (SON) that is trained to map each local patch to a vectorial descriptor that is sensitive to the characteristics (e.g. shape) of a central object, while being invariant to the variability of all other surrounding elements. Given a SON, a local image patch can be matched to a gallery of isolated elements using their SON-descriptors, thus producing a hypothesis about the shape of the central element in that patch. The image-level optimization based on integer programming can then pick a subset of the hypotheses to explain (parse) the whole image and disentangle groups of organisms. While sharing many similarities with existing analysis-by-synthesis approaches, our method avoids the need for stochastic search in the high-dimensional configuration space and numerous rendering operations at test-time. We show that our approach can parse microscopy images of three popular model organisms (the C.Elegans roundworms, the Drosophila larvae, and the E.Coli bacteria) even under significant crowding and overlaps between organisms. We speculate that the overall approach is applicable to a wider class of image parsing problems concerned with crowded articulated objects, for which rendering training images is possible.

To compress large datasets of high-dimensional descriptors, modern quantization schemes learn multiple codebooks and then represent individual descriptors as combinations of codewords. Once the codebooks are learned, these schemes encode descriptors independently. In contrast to that, we present a new coding scheme that arranges dataset descriptors into a set of arborescence graphs, and then encodes non-root descriptors by quantizing their displacements with respect to their parent nodes. By optimizing the structure of arborescences, our coding scheme can decrease the quantization error considerably, while incurring only minimal overhead on the memory footprint and the speed of nearest neighbor search in the compressed dataset compared to the independent quantization. The advantage of the proposed scheme is demonstrated in a series of experiments with datasets of SIFT and deep descriptors.

We present a new deep learning architecture (called Kdnetwork) that is designed for 3D model recognition tasks and works with unstructured point clouds. The new architecture performs multiplicative transformations and shares parameters of these transformations according to the subdivisions of the point clouds imposed onto them by kdtrees. Unlike the currently dominant convolutional architectures that usually require rasterization on uniform twodimensional or three-dimensional grids, Kd-networks do not rely on such grids in any way and therefore avoid poor scaling behavior. In a series of experiments with popular shape recognition benchmarks, Kd-networks demonstrate competitive performance in a number of shape recognition tasks such as shape classification, shape retrieval and shape part segmentation.

We revisit the idea of brain damage, i.e. the pruning of the coefficients of a neural network, and suggest how brain damage can be modified and used to speedup convolutional layers in ConvNets. The approach uses the fact that many efficient implementations reduce generalized convolutions to matrix multiplications. The suggested brain damage process prunes the convolutional kernel tensor in a group-wise fashion. After such pruning, convolutions can be reduced to multiplications of thinned dense matrices, which leads to speedup. We investigate different ways to add group-wise prunning to the learning process, and show that severalfold speedups of convolutional layers can be attained using group-sparsity regularizers. Our approach can adjust the shapes of the receptive fields in the convolutional layers, and even prune excessive feature maps from ConvNets, all in data-driven way.

Existing billion-scale nearest neighbor search systems have mostly been compared on a single dataset of a billion of SIFT vectors, where systems based on the Inverted Multi-Index (IMI) have been performing very well, achieving state-of-the-art recall in several milliseconds. SIFT-like descriptors, however, are quickly being replaced with descriptors based on deep neural networks (DNN) that provide better performance for many computer vision tasks. In this paper, we introduce a new dataset of one billion descriptors based on DNNs and reveal the relative inefficiency of IMI-based indexing for such descriptors compared to SIFT data. We then introduce two new indexing structures, the Non-Orthogonal Inverted Multi-Index (NO-IMI) and the Generalized Non-Orthogonal Inverted Multi-Index (GNO-IMI). We show that due to additional flexibility, the new structures are able to adapt to DNN descriptor distribution in a better way. In particular, extensive experiments on the new dataset demonstrate that these data structures provide considerably better trade-off between the speed of retrieval and recall, given similar amount of memory, as compared to the standard Inverted Multi-Index.

Gatys et al. recently demonstrated that deep networks can generate beautiful textures and stylized images from a single texture example. However, their methods requires a slow and memory-consuming optimization process. We propose here an alternative approach that moves the computational burden to a learning stage. Given a single example of a texture, our approach trains compact feed-forward convolutional networks to generate multiple samples of the same texture of arbitrary size and to transfer artistic style from a given image to any other image. The resulting networks are remarkably light-weight and can generate textures of quality comparable to Gatys et al., but hundreds of times faster. More generally, our approach highlights the power and flexibility of generative feed-forward models trained with complex and expressive loss functions.

We propose a new approach to designing visual markers (analogous to QR-codes, markers for augmented reality, and robotic fiducial tags) based on the advances in deep generative networks. In our approach, the markers are obtained as color images synthesized by a deep network from input bit strings, whereas another deep network is trained to recover the bit strings back from the photos of these markers. The two networks are trained simultaneously in a joint backpropagation process that takes characteristic photometric and geometric distortions associated with marker fabrication and capture into account. Additionally, a stylization loss based on statistics of activations in a pretrained classification network can be inserted into the learning in order to shift the marker appearance towards some texture prototype. In the experiments, we demonstrate that the markers obtained using our approach are capable of retaining bit strings that are long enough to be practical. The ability to automatically adapt markers according to the usage scenario and the desired capacity as well as the ability to combine information encoding with artistic stylization are the unique properties of our approach. As a byproduct, our approach provides an insight on the structure of patterns that are most suitable for recognition by ConvNets and on their ability to distinguish composite patterns.

We suggest a new loss for learning deep embeddings. The key characteristics of the new loss is the absence of tunable parameters and very good results obtained across a range of datasets and problems. The loss is computed by estimating two distribution of similarities for positive (matching) and negative (non-matching) point pairs, and then computing the probability of a positive pair to have a lower similarity score than a negative pair based on these probability estimates. We show that these operations can be performed in a simple and piecewise-differentiable manner using 1D histograms with soft assignment operations. This makes the proposed loss suitable for learning deep embeddings using stochastic optimization. The experiments reveal favourable results compared to recently proposed loss functions.

We introduce a new representation learning approach for domain adaptation, in which data at training and test time come from similar but different distributions. Our approach is directly inspired by the theory on domain adaptation suggesting that, for effective domain transfer to be achieved, predictions must be made based on features that cannot discriminate between the training (source) and test (target) domains. The approach implements this idea in the context of neural network architectures that are trained on labeled data from the source domain and unlabeled data from the target domain (no labeled target-domain data is necessary). As the training progresses, the approach promotes the emergence of features that are (i) discriminative for the main learning task on the source domain and (ii) indiscriminate with respect to the shift between the domains. We show that this adaptation behaviour can be achieved in almost any feed-forward model by augmenting it with few standard layers and a new gradient reversal layer. The resulting augmented architecture can be trained using standard backpropagation and stochastic gradient descent, and can thus be implemented with little effort using any of the deep learning packages. We demonstrate the success of our approach for two distinct classification problems (document sentiment analysis and image classification), where state-of-the-art domain adaptation performance on standard benchmarks is achieved. We also validate the approach for descriptor learning task in the context of person re-identification application.

We propose a new vector encoding scheme (tree quantization) that obtains lossy compact codes for high-dimensional vectors via tree-based dynamic programming. Similarly to several previous schemes such as product quantization, these codes correspond to codeword numbers within multiple codebooks. We propose an integer programming-based optimization that jointly recovers the coding tree structure and the codebooks by minimizing the compression error on a training dataset. In the experiments with diverse visual descriptors (SIFT, neural codes, Fisher vectors), tree quantization is shown to combine fast encoding and state-of-the-art accuracy in terms of the compression error, the retrieval performance, and the image classification error.

We revisit the well-known problem of gaze correction and present a solution based on supervised machine learning. At training time, our system observes pairs of images, where each pair contains the face of the same person with a fixed angular difference in gaze direction. It then learns to synthesize the second image of a pair from the first one. After learning, the system gets the ability to redirect the gaze of a previously unseen person by the same angular difference as in the training set. Unlike many previous solutions to gaze problem in videoconferencing, ours is purely monocular, i.e. it does not require any hardware apart from an in-built web-camera of a laptop. Being based on efficient machine learning predictors such as decision forests, the system is fast (runs in real-time on a single core of a modern laptop). In the paper, we demonstrate results on a variety of videoconferencing frames and evaluate the method quantitatively on the hold-out set of registered images. The supplementary video shows example sessions of our system at work.

Several recent works have shown that image descriptors produced by deep convolutional neural networks provide state-of-the-art performance for image classification and retrieval problems. It also has been shown that the activations from the convolutional layers can be interpreted as local features describing particular image regions. These local features can be aggregated using aggregating methods developed for local features (e.g. Fisher vectors), thus providing new powerful global descriptor. In this paper we investigate possible ways to aggregate local deep features to produce compact descriptors for image retrieval. First, we show that deep features and traditional hand-engineered features have quite different distributions of pairwise similarities, hence existing aggregation methods have to be carefully re-evaluated. Such re-evaluation reveals that in contrast to shallow features, the simple aggregation method based on sum pooling provides the best performance for deep convolutional features. This method is efficient, has few parameters, and bears little risk of overfitting when e.g. learning the PCA matrix. In addition, we suggest a simple yet efficient query expansion scheme suitable for the proposed aggregation method. Overall, the new compact global descriptor improves the state-of-the-art on four common benchmarks considerably.

Top-performing deep architectures are trained on massive amounts of labeled data. In the absence of labeled data for a certain task, domain adaptation often provides an attractive option given that labeled data of similar nature but from a different domain (e.g. synthetic images) are available. Here, we propose a new approach to domain adaptation in deep architectures that can be trained on large amount of labeled data from the source domain and large amount of unlabeled data from the target domain (no labeled target-domain data is necessary). As the training progresses, the approach promotes the emergence of "deep" features that are (i) discriminative for the main learning task on the source domain and (ii) invariant with respect to the shift between the domains. We show that this adaptation behaviour can be achieved in almost any feed-forward model by augmenting it with few standard layers and a simple new gradient reversal layer. The resulting augmented architecture can be trained using standard backpropagation. Overall, the approach can be implemented with little effort using any of the deep-learning packages. The method performs very well in a series of image classification experiments, achieving adaptation effect in the presence of big domain shifts and outperforming previous state-of-the-art on Office datasets.

A new data structure for efficient similarity search in very large datasets of high-dimensional vectors is introduced. This structure called the inverted multi-index generalizes the inverted index idea by replacing the standard quantization within inverted indices with product quantization. For very similar retrieval complexity and pre-processing time, inverted multi-indices achieve a much denser subdivision of the search space compared to inverted indices, while retaining their memory efficiency. Our experiments with large datasets of SIFT and GIST vectors demonstrate that because of the denser subdivision, inverted multi-indices are able to return much shorter candidate lists with higher recall. Augmented with a suitable reranking procedure, multi-indices were able to significantly improve the speed of approximate nearest neighbor search on the dataset of 1 billion SIFT vectors compared to the best previously published systems, while achieving better recall and incurring only few percent of memory overhead.

We introduce a new compression scheme for high-dimensional vectors that approximates the vectors using sums of M codewords coming from M different codebooks. We show that the proposed scheme permits efficient distance and scalar product computations between compressed and uncompressed vectors. We further suggest vector encoding and codebook learning algorithms that can minimize the coding error within the proposed scheme. In the experiments, we demonstrate that the proposed compression can be used instead of or together with product quantization. Compared to product quantization and its optimized versions, the proposed compression approach leads to lower coding approximation errors, higher accuracy of approximate nearest neighbor search in the datasets of visual descriptors, and lower image classification error, whenever the classifiers are learned on or applied to compressed vectors.

The objective of this work is to detect all instances of a class (such as cells or people) in an image. The instances may be partially overlapping and clustered, and hence quite challenging for traditional detectors, which aim at localizing individual instances. Our approach is to propose a set of candidate regions, and then select regions based on optimizing a global classification score, subject to the constraint that the selected regions are non-overlapping. Our novel contribution is to extend standard object detection by introducing separate classes for tuples of objects into the detection process. For example, our detector can pick a region containing two or three object instances, while assigning such region an appropriate label. We show that this formulation can be learned within the structured output SVM framework, and that the inference in such model can be accomplished using dynamic programming on a tree structured region graph. Furthermore, the learning only requires weak annotations - a dot on each instance. The improvement resulting from the addition of the capability to detect tuples of objects is demonstrated on quite disparate data sets: fluorescence microscopy images and UCSD pedestrians.

We propose a new method for the task of fine-grained visual categorization. The method builds a model of the base-level category that can be fitted to images, producing high-quality foreground segmentation and mid-level part localizations. The model can be learnt from the typical datasets available for fine-grained categorization, where the only annotation provided is a loose bounding box around the instance (e.g. bird) in each image. Both segmentation and part localizations are then used to encode the image content into a highly-discriminative visual signature. The model is symbiotic in that part discovery/localization is helped by segmentation and, conversely, the segmentation is helped by the detection (e.g. part layout). Our model builds on top of the part-based object category detector of Felzenszwalb et al., and also on the powerful Grab Cut segmentation algorithm of Rother et al., and adds a simple spatial saliency coupling between them. In our evaluation, the model improves the categorization accuracy over the state-of-the-art. It also improves over what can be achieved with an analogous system that runs segmentation and part-localization independently.

While modern off-the-shelf OCR engines show particularly high accuracy on scanned text, text detection and recognition in natural images still remains a challenging problem. Here, we demonstrate that OCR engines can still perform well on this harder task as long as appropriate image binarization is applied to input photographs. For such binarization, we systematically evaluate the performance of 12 binarization methods as well as of a new binarization algorithm that we propose here. Our evaluation includes different metrics and uses established natural image text recognition benchmarks (ICDAR 2003 and ICDAR 2011). Our main finding is thus the fact that image binarization methods combined with additional filtering of generated connected components and off-the-shelf OCR engines can achieve state-of-the-art performance for end-to-end text understanding in natural images.

A new data structure for efficient similarity search in very large dataseis of high-dimensional vectors is introduced. This structure called the inverted multi-index generalizes the inverted index idea by replacing the standard quantization within inverted indices with product quantization. For very similar retrieval complexity and preprocessing time, inverted multi-indices achieve a much denser subdivision of the search space compared to inverted indices, while retaining their memory efficiency. Our experiments with large dataseis of SIFT and GIST vectors demonstrate that because of the denser subdivision, inverted multi-indices are able to return much shorter candidate lists with higher recall. Augmented with a suitable reranking procedure, multi-indices were able to improve the speed of approximate nearest neighbor search on the dataset of 1 billion SIFT vectors by an order of magnitude compared to the best previously published systems, while achieving better recall and incurring only few percent of memory overhead.

Hough transform-based methods for detecting multiple objects use nonmaxima suppression or mode seeking to locate and distinguish peaks in Hough images. Such postprocessing requires the tuning of many parameters and is often fragile, especially when objects are located spatially close to each other. In this paper, we develop a new probabilistic framework for object detection which is related to the Hough transform. It shares the simplicity and wide applicability of the Hough transform but, at the same time, bypasses the problem of multiple peak identification in Hough images and permits detection of multiple objects without invoking nonmaximum suppression heuristics. Our experiments demonstrate that this method results in a significant improvement in detection accuracy both for the classical task of straight line detection and for a more modern category-level (pedestrian) detection problem.

A large number of novel encodings for bag of visual words models have been proposed in the past two years to improve on the standard histogram of quantized local features. Examples include locality-constrained linear encoding [23], improved Fisher encoding [17], super vector encoding [27], and kernel codebook encoding [20]. While several authors have reported very good results on the challenging PASCAL VOC classification data by means of these new techniques, differences in the feature computation and learning algorithms, missing details in the description of the methods, and different tuning of the various components, make it impossible to compare directly these methods and hard to reproduce the results reported. This paper addresses these shortcomings by carrying out a rigorous evaluation of these new techniques by: (1) fixing the other elements of the pipeline (features, learning, tuning); (2) disclosing all the implementation details, and (3) identifying both those aspects of each method which are particularly important to achieve good performance, and those aspects which are less critical. This allows a consistent comparative analysis of these encoding methods. Several conclusions drawn from our analysis cannot be inferred from the original publications.

The objective of this paper is the unsupervised segmentation of image training sets into foreground and background in order to improve image classification performance. To this end we introduce a new scalable, alternation-based algorithm for co-segmentation, BiCoS, which is simpler than many of its predecessors, and yet has superior performance on standard benchmark image datasets.

Graph cut optimization is one of the standard workhorses of image segmentation since for binary random field representations of the image, it gives globally optimal results and there are efficient polynomial time implementations. Often, the random field is applied over a flat partitioning of the image into non-intersecting elements, such as pixels or super-pixels. In the paper we show that if, instead of a flat partitioning, the image is represented by a hierarchical segmentation tree, then the resulting energy combining unary and boundary terms can still be optimized using graph cut (with all the corresponding benefits of global optimality and efficiency). As a result of such inference, the image gets partitioned into a set of segments that may come from different layers of the tree. We apply this formulation, which we call the pylon model, to the task of semantic segmentation where the goal is to separate an image into areas belonging to different semantic classes. The experiments highlight the advantage of inference on a segmentation tree (over a flat partitioning) and demonstrate that the optimization in the pylon model is able to flexibly choose the level of segmentation across the image. Overall, the proposed system has superior segmentation accuracy on several datasets (Graz-02, Stanford background) compared to previously suggested approaches.

Abstract—The paper introduces Hough forests, which are random forests adapted to perform a generalized Hough transform in an efficient way. Compared to previous Hough-based systems such as implicit shape models, Hough forests improve the performance of the generalized Hough transform for object detection on a categorical level. At the same time, their flexibility permits extensions of the Hough transform to new domains such as object tracking and action recognition. Hough forests can be regarded as task-adapted codebooks of local appearance that allow fast supervised training and fast matching at test time. They achieve high detection accuracy since the entries of such codebooks are optimized to cast Hough votes with small variance and since their efficiency permits dense sampling of local image patches or video cuboids during detection. The efficacy of Hough forests for a set of computer vision tasks is validated through experiments on a large set of publicly available benchmark data sets and comparisons with the state-of-the-art.

To detect multiple objects of interest, the methods based on Hough transform use non-maxima supression or mode seeking in order to locate and to distinguish peaks in Hough images. Such postprocessing requires tuning of extra parameters and is often fragile, especially when objects of interest tend to be closely located. In the paper, we develop a new probabilistic framework that is in many ways related to Hough transform, sharing its simplicity and wide applicability. At the same time, the framework bypasses the problem of multiple peaks identification in Hough images, and permits detection of multiple objects without invoking nonmaximum suppression heuristics. As a result, the experiments demonstrate a significant improvement in detection accuracy both for the classical task of straight line detection and for a more modern category-level (pedestrian) detection problem.

A number of 3D shape reconstruction algorithms, in particular 3D image segmentation methods, produce their results in the form of binary volumes, where a binary value indicates whether a voxel is associated with the interior or the exterior. For visualization purpose, it is often desirable to convert a binary volume into a surface representation. Straightforward extraction of the median isosurfaces for binary volumes using the marching cubes algorithm, however, produces jaggy, visually unrealistic meshes. Therefore, similarly to some previous works, we suggest to precede the isosurface extraction by replacing the original binary volume with a new continuous-valued embedding function, so that the zero-isosurface of the embedding function is smooth but at the same time consistent with the original binary volume. In contrast to previous work, computing such an embedding function in our case permits imposing a higher-order smoothness on the embedding function and involves solving a convex optimization problem. We demonstrate that the resulting separating surfaces are smoother and of better visual quality than minimal area separating surfaces extracted by previous approaches to the problem. The code of the algorithm is publicly available.

We propose a new supervised learning framework for visual object counting tasks, such as estimating the number of cells in a microscopic image or the number of humans in surveillance video frames. We focus on the practically-attractive case when the training images are annotated with dots (one dot per object). Our goal is to accurately estimate the count. However, we evade the hard task of learning to detect and localize individual object instances. Instead, we cast the problem as that of estimating an image density whose integral over any image region gives the count of objects within that region. Learning to infer such density can be formulated as a minimization of a regularized risk quadratic cost function. We introduce a new loss function, which is well-suited for such learning, and at the same time can be computed efficiently via a maximum subarray algorithm. The learning can then be posed as a convex quadratic program solvable with cutting-plane optimization. The proposed framework is very flexible as it can accept any domain-specific visual features. Once trained, our system provides accurate object counts and requires a very small time overhead over the feature extraction step, making it a good candidate for applications involving real-time processing or dealing with huge amount of visual data.

The efficient application of graph cuts to Markov Random Fields (MRFs) with multiple discrete or continuous labels remains an open question. In this paper, we demonstrate one possible way of achieving this by using graph cuts to combine pairs of suboptimal labelings or solutions. We call this combination process the fusion move. By employing recently developed graph-cut-based algorithms (so-called QPBO-graph cut), the fusion move can efficiently combine two proposal labelings in a theoretically sound way, which is in practice often globally optimal. We demonstrate that fusion moves generalize many previous graph-cut approaches, which allows them to be used as building blocks within a broader variety of optimization schemes than were considered before. In particular, we propose new optimization schemes for computer vision MRFs with applications to image restoration, stereo, and optical flow, among others. Within these schemes the fusion moves are used 1) for the parallelization of MRF optimization into several threads, 2) for fast MRF optimization by combining cheap-to-compute solutions, and 3) for the optimization of highly nonconvex continuous-labeled MRFs with 2D labels. Our final example is a nonvision MRF concerned with cartographic label placement, where fusion moves can be used to improve the performance of a standard inference method (loopy belief propagation).

User-provided object bounding box is a simple and popular interaction paradigm considered by many existing interactive image segmentation frameworks. However, these frameworks tend to exploit the provided bounding box merely to exclude its exterior from consideration and sometimes to initialize the energy minimization. In this paper, we discuss how the bounding box can be further used to impose a powerful topological prior, which prevents the solution from excessive shrinking and ensures that the user-provided box bounds the segmentation in a sufficiently tight way. The prior is expressed using hard constraints incorporated into the global energy minimization framework leading to an NP-hard integer program. We then investigate the possible optimization strategies including linear relaxation as well as a new graph cut algorithm called pinpointing. The latter can be used either as a rounding method for the fractional LP solution, which is provably better than thresholding-based rounding, or as a fast standalone heuristic. We evaluate the proposed algorithms on a publicly available dataset, and demonstrate the practical benefits of the new prior both qualitatively and quantitatively.

Accurate estimation of optical flow is a challenging task, which often requires addressing difficult energy optimization problems. To solve them, most top-performing methods rely on continuous optimization algorithms. The modeling accuracy of the energy in this case is often traded for its tractability. This is in contrast to the related problem of narrow-baseline stereo matching, where the top-performing methods employ powerful discrete optimization algorithms such as graph cuts and message-passing to optimize highly non-convex energies. In this paper, we demonstrate how similar non-convex energies can be formulated and optimized discretely in the context of optical flow estimation. Starting with a set of candidate solutions that are produced by fast continuous flow estimation algorithms, the proposed method iteratively fuses these candidate solutions by the computation of minimum cuts on graphs. The obtained continuous-valued fusion result is then further improved using local gradient descent. Experimentally, we demonstrate that the proposed energy is an accurate model and that the proposed discrete-continuous optimization scheme not only finds lower energy solutions than traditional discrete or continuous optimization techniques, but also leads to flow estimates that outperform the current state-of-the-art.

We propose a global optimization framework for 3D shape reconstruction from sparse noisy 3D measurements frequently encountered in range scanning, sparse feature-based stereo, and shape-from-X. In contrast to earlier local or banded optimization methods for shape fitting, we compute global optimum in the whole volume removing dependence on initial guess and sensitivity to numerous local minima. Our global method is based on two main ideas. First, we suggest a new regularization functional with a data alignment term that maximizes the number of (weakly-oriented) data points contained by a surface while allowing for some measurement errors. Second, we propose a touch-expand algorithm for finding a minimum cut on a huge 3D grid using an automatically adjusted band. This overcomes prohibitively high memory cost of graph cuts when computing globally optimal surfaces at high-resolution. Our results for sparse or incomplete 3D data from laser scanning and passive multi-view stereo are robust to noise, outliers, missing parts, and varying sampling density.

Image-based object modeling has emerged as an important computer vision application. Typically, the process starts with the acquisition of the image views of an object. These views are registered within the global coordinate system using structure-and-motion techniques, while on the next step the geometric shape of an object is recovered using stereo and/or silhouette cues. This paper considers the final step, which creates the texture map for the recovered geometry model. The approach proposed in the paper naturally starts by backprojecting original views onto the obtained surface. A texture is then mosaiced from these back projections, whereas the quality of the mosaic is maximized within the process of Markov random field energy optimization. Finally, the residual seams between the mosaic components are removed via seam levelling procedure, which is similar to gradient-domain stitching techniques recently proposed for image editing. Unlike previous approaches to the same problem, intensity blending as well as image resampling are avoided on all stages of the process, which ensures that the resolution of the produced texture is essentially the same as that of the original views. Importantly, due to restriction to non-greedy energy optimization techniques, good results are produced even in the presence of significant errors on image registration and geometric estimation steps.

Many computer vision applications rely on the efficient optimization of challenging, so-called non-submodular, binary pairwise MRFs. A promising graph cut based approach for optimizing such MRFs known as "roof duality" was recently introduced into computer vision. We study two methods which extend this approach. First, we discuss an efficient implementation of the "probing" technique introduced recently by Bows et al. (2006). It simplifies the MRF while preserving the global optimum. Our code is 400-700 faster on some graphs than the implementation of the work of Bows et al. (2006). Second, we present a new technique which takes an arbitrary input labeling and tries to improve its energy. We give theoretical characterizations of local minima of this procedure. We applied both techniques to many applications, including image segmentation, new view synthesis, super-resolution, diagram recognition, parameter learning, texture restoration, and image deconvolution. For several applications we see that we are able to find the global minimum very efficiently, and considerably outperform the original roof duality approach. In comparison to existing techniques, such as graph cut, TRW, BP, ICM, and simulated annealing, we nearly always find a lower energy.

Markov Random Fields (MRFs) are ubiquitous in low- level computer vision. In this paper, we propose a new approach to the optimization of multi-labeled MRFs. Similarly to a-expansion it is based on iterative application of binary graph cut. However, the number of binary graph cuts required to compute a labelling grows only logarithmically with the size of label space, instead of linearly. We demonstrate that for applications such as optical flow, image restoration, and high resolution stereo, this gives an order of magnitude speed-up, for comparable energies. Iterations are performed by "fusion" of solutions, done with QPBO which is related to graph cut but can deal with non-submodularity. At convergence, the method achieves optima on a par with the best competitors, and sometimes even exceeds them.

Our work was inspired by recent advances in image segmentation where fluxbased functionals significantly improved alignment of object boundaries. We propose a novel photoflux functional for multi-view 3D reconstruction that is closely related to properties of photohulls. Our photohull prior can be combined with regularization. Thus, this work unifies two major groups of multiview stereo techniques: “space carving ” and “deformable models”. Our approach combines benefits of both groups and allows to recover fine shape details without oversmoothing while robustly handling noise. Photoflux provides data-driven ballooning force that helps to segment thin structures or holes. Photoflux maximizing shapes can be also seen as regularized Laplacian zero-crossings [3]. We discuss several versions of photoflux functional based on global, local, or non-deterministic visibility models. Some forms of photoflux can be easily added into standard regularization techniques. For other forms we propose new optimization methods. 1

Spatial patches have been introduced (Ivanov and Kuzmin, 2001) as an alternative primitive for the representation of 3D objects potentially acquired from real life. Defined as range images of small regions of an object's surface, they have proven to allow for efficient storage and rendering of such surfaces, and they also exhibit great descriptive capabilities. We present several technologies that we have developed for generating 3D objects represented by spatial patches. These technologies allowed us to convert explicitly defined surfaces, textured polygonal meshes, range images and point clouds to the proposed representation. We also describe how advantageous properties of the new primitive are exploited for efficient reconstruction from range images, discuss the particular problems that we faced during our research, provide many examples to demonstrate the results that we achieved, and conclude with discussion of future work.