Level-5 driving automation requires a robust visual perception system that can parse input images under any condition. However, existing driving datasets for dense semantic perception are either dominated by images captured under normal conditions or are small in scale. To address this, we introduce ACDC, the Adverse Conditions Dataset with Correspondences for training and testing methods for diverse semantic perception tasks on adverse visual conditions. ACDC consists of a large set of 8012 images, half of which (4006) are equally distributed between four common adverse conditions: fog, nighttime, rain, and snow. Each adverse-condition image comes with a high-quality pixel-level panoptic annotation, a corresponding image of the same scene under normal conditions, and a binary mask that distinguishes between intra-image regions of clear and uncertain semantic content. 1503 of the corresponding normal-condition images feature panoptic annotations, raising the total annotated images to 5509. ACDC supports the standard tasks of semantic segmentation, object detection, instance segmentation, and panoptic segmentation, as well as the newly introduced uncertainty-aware semantic segmentation. A detailed empirical study demonstrates the challenges that the adverse domains of ACDC pose to state-of-the-art supervised and unsupervised approaches and indicates the value of our dataset in steering future progress in the field.

As perception models continue to develop, the need for large-scale datasets increases. However, data annotation remains far too expensive to effectively scale and meet the demand. Synthetic datasets provide a solution to boost model performance with substantially reduced costs. However, current synthetic datasets remain limited in their scope, realism, and are designed for specific tasks and applications. In this work, we present RealDriveSim, a realistic multi-modal synthetic dataset for autonomous driving that not only supports popular 2D computer vision applications but also their LiDAR counterparts, providing fine-grained annotations for up to 64 classes. We extensively evaluate our dataset for a wide range of applications and domains, demonstrating state-of-the-art results compared to existing synthetic benchmarks. The dataset is publicly available at https//zrealdrivesim.github.io/.

The progress on Hyperspectral image (HSI) super-resolution (SR) is still lagging behind the research of RGB image SR. HSIs usually have a high number of spectral bands, so accurately modeling spectral band interaction for HSI SR is hard. Also, training data for HSI SR is hard to obtain so the dataset is usually rather small. In this work, we propose a new test-time training method to tackle this problem. Specifically, a novel self-training framework is developed, where more accurate pseudo-labels and more accurate LR-HR relationships are generated so that the model can be further trained with them to improve performance. In order to better support our test-time training method, we also propose a new network architecture to learn HSI SR without modeling spectral band interaction and propose a new data augmentation method Spectral Mixup to increase the diversity of the training data at test time. We also collect a new HSI dataset with a diverse set of images of interesting objects ranging from food to vegetation, to materials, and to general scenes. Extensive experiments on multiple datasets show that our method can improve the performance of pre-trained models significantly after test-time training and outperform competing methods significantly for HSI SR.

Instance segmentation of point clouds is a crucial task in 3D field with numerous applications that involve localizing and segmenting objects in a scene. However, achieving sat-isfactory results requires a large number of manual annotations, which is time-consuming and expensive. To alleviate dependency on annotations, we propose a novelframework, FreePoint, for underexplored unsupervised class-agnostic instance segmentation on point clouds. In detail, we represent the point features by combining coordinates, colors, and self-supervised deep features. Based on the point features, we perform a bottom-up multicut algorithm to seg-ment point clouds into coarse instance masks as pseudo labels, which are used to train a point cloud instance segmen-tation model. We propose an id-as-feature strategy at this stage to alleviate the randomness of the multicut algorithm and improve the pseudo labels' quality. During training, we propose a weakly-supervised two-step training strategy and corresponding losses to overcome the inaccuracy of coarse masks. FreePoint has achieved breakthroughs in un-supervised class-agnostic instance segmentation on point clouds and outperformed previous traditional methods by over 18.2% and a competitive concurrent work UnScene3D by 5.5% in AP. Additionally, when used as a pretext task and fine-tuned on S3DIS, FreePoint performs significantly better than existing self-supervised pre-training methods with limited annotations and surpasses CSC by 6.0% in AP with 10% annotation masks. Code will be released at https://github.com/zzk273/FreePoint.

Event cameras are gaining popularity due to their unique properties, such as their low latency and high dynamic range. One task where these benefits can be crucial is real-time object detection. However, RGB detectors still outperform event-based detectors due to the sparsity of the event data and missing visual details. In this paper, we propose a cross-modality feature distillation method that can focus on regions where the knowledge distillation works best to shrink the detection performance gap between these two modalities. We achieve this by using an object-centric slot attention mechanism that can iteratively decouple feature maps into object-centric features and corresponding pixel-features used for distillation. We evaluate our novel distillation approach on a synthetic and a real event dataset with aligned grayscale images as a teacher modality. We show that object-centric distillation allows to significantly improve the performance of the event-based student object detector, nearly halving the performance gap with respect to the teacher.

Efficient relocalization is essential for intelligent vehicles when GPS reception is insufficient or sensor-based localization fails. Recent advances in Bird’s-Eye-View (BEV) segmentation allow for accurate estimation of local scene appearance and in turn, can benefit the relocalization of the vehicle. However, one downside of BEV methods is the heavy computation required to leverage the geometric constraints. This paper presents U-BEV, a U-Net inspired architecture that extends the current state-of-the-art by allowing the BEV to reason about the scene on multiple height layers before flattening the BEV features. We show that this extension boosts the performance of the U-BEV by up to 4.11 IoU. Additionally, we combine the encoded neural BEV with a differentiable template matcher to perform relocalization on neural SD-map data. The model is fully end-to-end trainable and outperforms transformer-based BEV methods of similar computational complexity by 1.7 to 2.8 mIoU and BEV-based relocalization by over 26% Recall Accuracy on the nuScenes dataset.

As 3D perception problems grow in popularity and the need for large-scale labeled datasets for LiDAR semantic segmentation increase, new methods arise that aim to reduce the necessity for dense annotations by employing weakly-supervised training. However these methods continue to show weak boundary estimation and high false negative rates for small objects and distant sparse regions. We argue that such weaknesses can be compensated by using RGB images which provide a denser representation of the scene. We propose an image-guidance network (IGNet) which builds upon the idea of distilling high level feature information from a domain adapted synthetically trained 2D semantic segmentation network. We further utilize a one-way contrastive learning scheme alongside a novel mixing strategy called FOVMix, to combat the horizontal field-of-view mismatch between the two sensors and enhance the effects of image guidance. IGNet achieves state-of-the-art results for weakly-supervised LiDAR semantic segmentation on ScribbleKITTI, boasting up to 98% relative performance to fully supervised training with only 8% labeled points, while introducing no additional annotation burden or computational/memory cost during inference. Furthermore, we show that our contributions also prove effective for semi-supervised training, where IGNet claims state-of-the-art results on both ScribbleKITTI and SemanticKITTI.

Unsupervised domain adaptation (UDA) and domain generalization (DG) enable machine learning models trained on a source domain to perform well on unlabeled or even unseen target domains. As previous UDA&DG semantic segmentation methods are mostly based on outdated networks, we benchmark more recent architectures, reveal the potential of Transformers, and design the DAFormer network tailored for UDA&DG. It is enabled by three training strategies to avoid overfitting to the source domain: While (1) Rare Class Sampling mitigates the bias toward common source domain classes, (2) a Thing-Class ImageNet Feature Distance and (3) a learning rate warmup promote feature transfer from ImageNet pretraining. As UDA&DG are usually GPU memory intensive, most previous methods downscale or crop images. However, low-resolution predictions often fail to preserve fine details while models trained with cropped images fall short in capturing long-range, domain-robust context information. Therefore, we propose HRDA, a multi-resolution framework for UDA&DG, that combines the strengths of small high-resolution crops to preserve fine segmentation details and large low-resolution crops to capture long-range context dependencies with a learned scale attention. DAFormer and HRDA significantly improve the state-of-the-art UDA&DG by more than 10 mIoU on 5 different benchmarks.

Motion prediction is crucial for autonomous driving systems to understand complex driving scenarios and make informed decisions. However, this task is challenging due to the diverse behaviors of traffic participants and complex environmental contexts. In this paper, we propose Motion TRansformer (MTR) frameworks to address these challenges. The initial MTR framework utilizes a transformer encoder-decoder structure with learnable intention queries, enabling efficient and accurate prediction of future trajectories. By customizing intention queries for distinct motion modalities, MTR improves multimodal motion prediction while reducing reliance on dense goal candidates. The framework comprises two essential processes: global intention localization, identifying the agent's intent to enhance overall efficiency, and local movement refinement, adaptively refining predicted trajectories for improved accuracy. Moreover, we introduce an advanced MTR++ framework, extending the capability of MTR to simultaneously predict multimodal motion for multiple agents. MTR++ incorporates symmetric context modeling and mutually-guided intention querying modules to facilitate future behavior interaction among multiple agents, resulting in scene-compliant future trajectories. Extensive experimental results demonstrate that the MTR framework achieves state-of-the-art performance on the highly-competitive motion prediction benchmarks, while the MTR++ framework surpasses its precursor, exhibiting enhanced performance and efficiency in predicting accurate multimodal future trajectories for multiple agents.

There has been a lot of effort in improving the performance of unsupervised domain adaptation for semantic segmentation task, however, there is still a huge gap in performance when compared with supervised learning. In this work, we propose a common framework to use different weak labels, e.g., image, point and coarse labels from the target domain to reduce this performance gap. Specifically, we propose to learn better prototypes that are representative class features by exploiting these weak labels. We use these improved prototypes for the contrastive alignment of class features. In particular, we perform two different feature alignments: first, we align pixel features with proto-types within each domain and second, we align pixel features from the source to prototype of target domain in an asymmetric way. This asymmetric alignment is beneficial as it preserves the target features during training, which is essential when weak labels are available from the target domain. Our experiments on various benchmarks show that our framework achieves significant improvement compared to existing works and can reduce the performance gap with supervised learning. Code will be available at https://github.com/anurag-198/WDASS.

Current semantic segmentation models have achieved great success under the independent and identically distributed (i.i.d.) condition. However, in real-world applications, test data might come from a different domain than training data. Therefore, it is important to improve model robustness against domain differences. This work studies semantic segmentation under the domain generalization setting, where a model is trained only on the source domain and tested on the unseen target domain. Existing works show that Vision Transformers are more robust than CNNs and show that this is related to the visual grouping property of self-attention. In this work, we propose a novel hierarchical grouping transformer (HGFormer) to explicitly group pixels to form part-level masks and then whole-level masks. The masks at different scales aim to segment out both parts and a whole of classes. HGFormer combines mask classification results at both scales for class label prediction. We assemble multiple interesting cross-domain settings by using seven public semantic segmentation datasets. Experiments show that HGFormer yields more robust semantic segmentation results than per-pixel classification methods and flat-grouping transformers, and outperforms previous methods significantly. Code will be available at https://github.com/dingjianswl0l/HGFormer.

Federated learning-based semantic segmentation (FSS) has drawn widespread attention via decentralized training on local clients. However, most FSS models assume categories are fixed in advance, thus heavily undergoing forgetting on old categories in practical applications where local clients receive new categories incrementally while have no memory storage to access old classes. Moreover, new clients collecting novel classes may join in the global training of FSS, which further exacerbates catastrophic forgetting. To surmount the above challenges, we propose a Forgetting-Balanced Learning (FBL) model to address heterogeneous forgetting on old classes from both intra-client and interclient aspects. Specifically, under the guidance of pseudo labels generated via adaptive class-balanced pseudo labeling, we develop a forgetting-balanced semantic compensation loss and a forgetting-balanced relation consistency loss to rectify intra-client heterogeneous forgetting of old categories with background shift. It performs balanced gradient propagation and relation consistency distillation within local clients. Moreover, to tackle heterogeneous forgetting from inter-client aspect, we propose a task transition monitor. It can identify new classes under privacy protection and store the latest old global model for relation distillation. Qualitative experiments reveal large improvement of our model against comparison methods. The code is available at https://github.com/JiahuaDong/FISS.

Unsupervised domain adaptation (UDA) for semantic segmentation aims at improving the model performance on the unlabeled target domain by leveraging a labeled source domain. Existing approaches have achieved impressive progress by utilizing pseudo-labels on the unlabeled target-domain images. Yet the low-quality pseudo-labels, arising from the domain discrepancy, inevitably hinder the adaptation. This calls for effective and accurate approaches to estimating the reliability of the pseudo-labels, in order to rectify them. In this paper, we propose to estimate the rectification values of the predicted pseudo-labels with implicit neural representations. We view the rectification value as a signal defined over the continuous spatial domain. Taking an image coordinate and the nearby deep features as inputs, the rectification value at a given coordinate is predicted as an output. This allows us to achieve high-resolution and detailed rectification values estimation, important for accurate pseudo-label generation at mask boundaries in particular. The rectified pseudo-labels are then leveraged in our rectification-aware mixture model (RMM) to be learned end-to-end and help the adaptation. We demonstrate the effectiveness of our approach on different UDA benchmarks, including synthetic-to-real and day-to-night. Our approach achieves superior results compared to state-of-the-art. The implementation is available at https://github.com/ETHRuiGong/IR2F.

In unsupervised domain adaptation (UDA), a model trained on source data (e.g. synthetic) is adapted to target data (e.g. real-world) without access to target annotation. Most previous UDA methods struggle with classes that have a similar visual appearance on the target domain as no ground truth is available to learn the slight appearance differences. To address this problem, we propose a Masked Image Consistency (MIC) module to enhance UDA by learning spatial context relations of the target domain as additional clues for robust visual recognition. MIC enforces the consistency between predictions of masked target images, where random patches are withheld, and pseudo-labels that are generated based on the complete image by an exponential moving average teacher. To minimize the consistency loss, the network has to learn to infer the predictions of the masked regions from their context. Due to its simple and universal concept, MIC can be integrated into various UDA methods across different visual recognition tasks such as image classification, semantic segmentation, and object detection. MIC significantly improves the state-of-the-art performance across the different recognition tasks for synthetic-to-real, day-to-nighttime, and clear-toadverse-weather UDA. For instance, MIC achieves an un-precedented UDA performance of 75.9 mIoU and 92.8% on GTA→Cityscapes and VisDA-2017, respectively, which corresponds to an improvement of +2.1 and +3.0 percent points over the previous state of the art. The implementation is available at https://github.com/lhoyer/MIC.

Masked signal modeling has greatly advanced self-supervised pre-training for language and 2D images. However, it is still not fully explored in 3D scene understanding. Thus, this paper introduces Masked Shape Prediction (MSP), a new framework to conduct masked signal modeling in 3D scenes. MSP uses the essential 3D semantic cue, i.e., geometric shape, as the prediction target for masked points. The context-enhanced shape target consisting of explicit shape context and implicit deep shape feature is proposed to facilitate exploiting contextual cues in shape prediction. Meanwhile, the pre-training architecture in MSP is carefully designed to alleviate the masked shape leakage from point coordinates. Experiments on multiple 3D understanding tasks on both indoor and outdoor datasets demon-strate the effectiveness of MSP in learning good feature representations to consistently boost downstream performance.

With autonomous industries on the rise, domain adaptation of the visual perception stack is an important research direction due to the cost savings promise. Much prior art was dedicated to domain-adaptive semantic segmentation in the synthetic-to-real context. Despite being a crucial output of the perception stack, panoptic segmentation has been largely overlooked by the domain adaptation community. Therefore, we revisit well-performing domain adaptation strategies from other fields, adapt them to panoptic segmentation, and show that they can effectively enhance panoptic domain adaptation. Further, we study the panoptic network design and propose a novel architecture (EDAPS) designed explicitly for domain-adaptive panoptic segmentation. It uses a shared, domain-robust transformer encoder to facilitate the joint adaptation of semantic and instance features, but task-specific decoders tailored for the specific requirements of both domain-adaptive semantic and instance segmentation. As a result, the performance gap seen in challenging panoptic benchmarks is substantially narrowed. EDAPS significantly improves the state-of-the-art performance for panoptic segmentation UDA by a large margin of 20% on SYNTHIA-to-Cityscapes and even 72% on the more challenging SYNTHIA-to-Mapillary Vistas. The implementation is available at https://github.com/susaha/edaps.

Semi-supervised learning (SSL) methods effectively leverage unlabeled data to improve model generalization. However, SSL models often underperform in open-set scenarios, where unlabeled data contain outliers from novel categories that do not appear in the labeled set. In this paper, we study the challenging and realistic open-set SSL setting, where the goal is to both correctly classify inliers and to detect outliers. Intuitively, the inlier classifier should be trained on inlier data only. However, we find that inlier classification performance can be largely improved by incorporating high-confidence pseudo-labeled data, regardless of whether they are inliers or outliers. Also, we propose to utilize non-linear transformations to separate the features used for inlier classification and outlier detection in the multi-task learning framework, preventing adverse effects between them. Additionally, we introduce pseudo-negative mining, which further boosts outlier detection performance. The three ingredients lead to what we call Simple but Strong Baseline (SSB) for open-set SSL. In experiments, SSB greatly improves both inlier classification and outlier detection performance, outperforming existing methods by a large margin. Our code will be released at https://github.com/YUE-FAN/SSB.

Test-time domain adaptation, i.e. adapting source-pretrained models to the test data on-the-fly in a source-free, unsupervised manner, is a highly practical yet very challenging task. Due to the domain gap between source and target data, inference quality on the target domain can drop drastically especially in terms of absolute scale of depth. In addition, unsupervised adaptation can degrade the model performance due to inaccurate pseudo labels. Furthermore, the model can suffer from catastrophic forgetting when errors are accumulated over time. We propose a test-time domain adaptation framework for monocular depth estimation which achieves both stability and adaptation performance by benefiting from both self-training of the supervised branch and pseudo labels from self-supervised branch, and is able to tackle the above problems: our scale alignment scheme aligns the input features between source and target data, correcting the absolute scale inference on the target domain; with pseudo label consistency check, we select confident pixels thus improve pseudo label quality; regularisation and self-training schemes are applied to help avoid catastrophic forgetting. Without requirement of further supervisions on the target domain, our method adapts the source-trained models to the test data with significant improvements over the direct inference results, providing scale-aware depth map outputs that outperform the state-of-the-arts. Code is available at https://github.com/Malefikus/ada-depth.

Data-driven simulation has become a favorable way to train and test autonomous driving algorithms. The idea of replacing the actual environment with a learned simulator has also been explored in model-based reinforcement learning in the context of world models. In this work, we show data-driven traffic simulation can be formulated as a world model. We present TrafficBots, a multi-agent policy built upon motion prediction and end-to-end driving, and based on TrafficBots we obtain a world model tailored for the planning module of autonomous vehicles. Existing data-driven traffic simulators are lacking configurability and scalability. To generate configurable behaviors, for each agent we introduce a destination as nav-igational information, and a time-invariant latent personality that specifies the behavioral style. To improve the scalability, we present a new scheme of positional encoding for angles, allowing all agents to share the same vectorized context and the use of an architecture based on dot-product attention. As a result, we can simulate all traffic participants seen in dense urban scenarios. Experiments on the Waymo open motion dataset show TrafficBots can simulate realistic multi-agent behaviors and achieve good performance on the motion prediction task.

Depth estimation is one of the essential tasks to be addressed when creating mobile autonomous systems. While monocular depth estimation methods have improved in recent times, depth completion provides more accurate and reliable depth maps by additionally using sparse depth information from other sensors such as LiDAR. However, current methods are specifically trained for a single LiDAR sensor. As the scanning pattern differs between sensors, every new sensor would require re- training a specialized depth completion model, which is computationally inefficient and not flexible. Therefore, we propose to dynamically adapt the depth completion model to the used sensor type enabling LiDAR adaptive depth completion. Specifically, we propose a meta depth completion network that uses data patterns derived from the data to learn a task network to alter weights of the main depth completion network to solve a given depth completion task effectively. The method demonstrates a strong capability to work on multiple LiDAR scanning patterns and can also generalize to scanning patterns that are unseen during training. While using a single model, our method yields significantly better results than a non-adaptive baseline trained on different LiDAR patterns. It outperforms LiDAR-specific expert models for very sparse cases. These advantages allow flexible deployment of a single depth completion model on different sensors, which could also prove valuable to process the input of nascent LiDAR technology with adaptive instead of fixed scanning patterns. The source code is available at github.com/wbkit/ResLAN

A single-shot multispectral camera equipped with an optimized color filter array (CFA) has the potential to deliver a fast and low-cost hyperspectral (HS) imaging system. Previous solutions are largely restricted to designing demosaicing algorithms for fixed CFAs – be it the Bayer color pattern or evenly-spaced spectral multiplexing patterns. Since sampling and reconstruction are tightly-coupled, the ability to search for an optimal solution is severely constrained by using predefined CFAs. In this work, we simultaneously address the problem of spectral band selection, CFA design, image demosaicing, and spectral image recovery in a joint learning framework for single-shot HS imaging. We propose a reinforcement learning (RL) based method for spectral band selection and a novel neural network for CFA generation, image demosaicing, and HS image recovery. The final spectral reconstruction accuracy is used to supervise the training of the main network to maximize the synergies between those tightly-related tasks. The RL method regards the main network as an agent to collect reward. Our final method delivers a simple setup – as simple as an RGB camera – for HS imaging. Experimental results show that our method outperforms competing methods by a large margin.

Abstract Training deep networks for semantic segmentation requires large amounts of labeled training data, which presents a major challenge in practice, as labeling segmentation masks is a highly labor-intensive process. To address this issue, we present a framework for semi-supervised and domain-adaptive semantic segmentation, which is enhanced by self-supervised monocular depth estimation (SDE) trained only on unlabeled image sequences. In particular, we utilize SDE as an auxiliary task comprehensively across the entire learning framework: First, we automatically select the most useful samples to be annotated for semantic segmentation based on the correlation of sample diversity and difficulty between SDE and semantic segmentation. Second, we implement a strong data augmentation by mixing images and labels using the geometry of the scene. Third, we transfer knowledge from features learned during SDE to semantic segmentation by means of transfer and multi-task learning. And fourth, we exploit additional labeled synthetic data with Cross-Domain DepthMix and Matching Geometry Sampling to align synthetic and real data. We validate the proposed model on the Cityscapes dataset, where all four contributions demonstrate significant performance gains, and achieve state-of-the-art results for semi-supervised semantic segmentation as well as for semi-supervised domain adaptation. In particular, with only 1/30 of the Cityscapes labels, our method achieves 92% of the fully-supervised baseline performance and even 97% when exploiting additional data from GTA. The source code is available at https://github.com/lhoyer/improving_segmentation_with_selfsupervised_depth .

Humans can robustly recognize and localize objects by using visual and/or auditory cues. While machines are able to do the same with visual data already, less work has been done with sounds. This work develops an approach for scene understanding purely based on binaural sounds. The considered tasks include predicting the semantic masks of sound-making objects, the motion of sound-making objects, and the depth map of the scene. To this aim, we propose a novel sensor setup and record a new audio-visual dataset of street scenes with eight professional binaural microphones and a 360$\mathrm{^{\circ }}$camera. The co-existence of visual and audio cues is leveraged for supervision transfer. In particular, we employ a cross-modal distillation framework that consists of multiple vision ‘teacher’ methods and a sound ‘student’ method – the student method is trained to generate the same results as the teacher methods do. This way, the auditory system can be trained without using human annotations. To further boost the performance, we propose another novel auxiliary task, coined Spatial Sound Super-Resolution, to increase the directional resolution of sounds. We then formulate the four tasks into one end-to-end trainable multi-tasking network aiming to boost the overall performance. Experimental results show that 1) our method achieves good results for all four tasks, 2) the four tasks are mutually beneficial – training them together achieves the best performance, 3) the number and orientation of microphones are both important, and 4) features learned from the standard spectrogram and features obtained by the classic signal processing pipeline are complementary for auditory perception tasks. The data and code are released on the project page:https://www.trace.ethz.ch/publications/2020/sound_perception/index.html.

Test-time domain adaptation aims to adapt a source pre-trained model to a target domain without using any source data. Existing works mainly consider the case where the target domain is static. However, real-world machine perception systems are running in non-stationary and continually changing environments where the target domain distribution can change over time. Existing methods, which are mostly based on self-training and entropy regularization, can suffer from these non-stationary environments. Due to the distribution shift over time in the target domain, pseudo-labels become unreliable. The noisy pseudo-labels can further lead to error accumulation and catastrophic forgetting. To tackle these issues, we propose a continual test-time adaptation approach (CoTTA) which comprises two parts. Firstly, we propose to reduce the error accumulation by using weight-averaged and augmentation-averaged predictions which are often more accurate. On the other hand, to avoid catastrophic forgetting, we propose to stochastically restore a small part of the neurons to the source pre-trained weights during each iteration to help preserve source knowledge in the longterm. The proposed method enables the longterm adaptation for all parameters in the network. CoTTA is easy to implement and can be readily incorporated in off-the-shelf pre-trained models. We demonstrate the effectiveness of our approach on four classification tasks and a segmentation task for continual test-time adaptation, on which we outperform existing methods. Our code is available at https://gin.ee/cotta.

We propose a pipeline to generate Neural Radiance Fields (NeRF) of an object or a scene of a specific class, conditioned on a single input image. This is a challenging task, as training NeRF requires multiple views of the same scene, coupled with corresponding poses, which are hard to obtain. Our method is based on$\pi$-GAN, a generative model for unconditional 3D-aware image synthesis, which maps random latent codes to radiance fields of a class of objects. We jointly optimize (1) the$\pi$-GAN objective to utilize its high-fidelity 3D-aware generation and (2) a carefully designed reconstruction objective. The latter includes an encoder coupled with$\pi$-GAN generator to form an autoencoder. Unlike previous few-shot NeRF approaches, our pipeline is unsupervised, capable of being trained with independent images without 3D, multi-view, or pose supervision. Applications of our pipeline include 3d avatar generation, object-centric novel view synthesis with a single input image, and 3d-aware super-resolution, to name a few.

Zero-shot semantic segmentation (ZS3) aims to segment the novel categories that have not been seen in the training. Existing works formulate ZS3 as a pixel-level zeroshot classification problem, and transfer semantic knowledge from seen classes to unseen ones with the help of language models pre-trained only with texts. While simple, the pixel-level ZS3 formulation shows the limited capability to integrate vision-language models that are often pre-trained with image-text pairs and currently demonstrate great potential for vision tasks. Inspired by the observation that humans often perform segment-level semantic labeling, we propose to decouple the ZS3 into two sub-tasks: 1) a classagnostic grouping task to group the pixels into segments. 2) a zero-shot classification task on segments. The former task does not involve category information and can be directly transferred to group pixels for unseen classes. The latter task performs at segment-level and provides a natural way to leverage large-scale vision-language models pre-trained with image-text pairs (e.g. CLIP) for ZS3. Based on the decoupling formulation, we propose a simple and effective zero-shot semantic segmentation model, called ZegFormer, which outperforms the previous methods on ZS3 standard benchmarks by large margins, e.g., 22 points on the PAS-CAL VOC and 3 points on the COCO-Stuff in terms of mIoU for unseen classes. Code will be released at https://github.com/dingjiansw101/ZegFormer.

Recently, crowd density estimation has received increasing attention. The main challenge for this task is to achieve high-quality manual annotations on a large amount of training data. To avoid reliance on such annotations, previous works apply unsupervised domain adaptation (UDA) techniques by transferring knowledge learned from easily accessible synthetic data to real-world datasets. However, current state-of-the-art methods either rely on external data for training an auxiliary task or apply an expensive coarse-to-fine estimation. In this work, we aim to develop a new adversarial learning based method, which is simple and efficient to apply. To reduce the domain gap between the synthetic and real data, we design a bi-level alignment framework (BLA) consisting of (1) task-driven data alignment and (2) fine-grained feature alignment. In contrast to previous domain augmentation methods, we introduce AutoML to search for an optimal transform on source, which well serves for the downstream task. On the other hand, we do fine-grained alignment for foreground and background separately to alleviate the alignment difficulty. We evaluate our approach on five real-world crowd counting benchmarks, where we outperform existing approaches by a large margin. Also, our approach is simple, easy to implement and efficient to apply. The code is publicly available at https://github.com/Yankeegsj/BLA.

3D object detection is a central task for applications such as autonomous driving, in which the system needs to localize and classify surrounding traffic agents, even in the presence of adverse weather. In this paper, we address the problem of LiDAR-based 3D object detection under snow-fall. Due to the difficulty of collecting and annotating training data in this setting, we propose a physically based method to simulate the effect of snowfall on real clear-weather LiDAR point clouds. Our method samples snow particles in 2D space for each LiDAR line and uses the in-duced geometry to modify the measurement for each LiDAR beam accordingly. Moreover, as snowfall often causes wet-ness on the ground, we also simulate ground wetness on LiDAR point clouds. We use our simulation to generate par-tially synthetic snowy LiDAR data and leverage these data for training 3D object detection models that are robust to snowfall. We conduct an extensive evaluation using several state-of-the-art 3D object detection methods and show that our simulation consistently yields significant performance gains on the real snowy STF dataset compared to clear-weather baselines and competing simulation approaches, while not sacrificing performance in clear weather. Our code is available at github.com/SysCV/LiDAR_snow_sim.

As acquiring pixel-wise annotations of real-world images for semantic segmentation is a costly process, a model can instead be trained with more accessible synthetic data and adapted to real images without requiring their annotations. This process is studied in unsupervised domain adaptation (UDA). Even though a large number of methods propose new adaptation strategies, they are mostly based on outdated network architectures. As the influence of recent network architectures has not been systematically studied, we first benchmark different network architectures for UDA and newly reveal the potential of Transformers for UDA semantic segmentation. Based on the findings, we propose a novel UDA method, DAFormer. The network architecture of DAFormer consists of a Transformer encoder and a multi-level context-aware feature fusion decoder. It is enabled by three simple but crucial training strategies to stabilize the training and to avoid overfitting to the source domain: While (1) Rare Class Sampling on the source domain improves the quality of the pseudo-labels by mitigating the confirmation bias of self-training toward common classes, (2) a Thing-Class ImageNet Feature Distance and (3) a learning rate warmup promote feature transfer from ImageNet pretraining. DAFormer represents a major advance in UDA. It improves the state of the art by 10.8 mIoU for GTA→Cityscapes and 5.4 mIoU for Syruhia→Cityscapes and enables learning even difficult classes such as train, bus, and truck well. The implementation is available at https://github.com/lhoyer/DAFormer.

Although considerable progress has been made in semantic scene understanding under clear weather, it is still a tough problem under adverse weather conditions, such as dense fog, due to the uncertainty caused by imperfect observations. Besides, difficulties in collecting and labeling foggy images hinder the progress of this field. Considering the success in semantic scene understanding under clear weather, we think it is reasonable to transfer knowledge learned from clear images to the foggy domain. As such, the problem becomes to bridge the domain gap between clear images and foggy images. Unlike previous methods that mainly focus on closing the domain gap caused by fog - defogging the foggy images or fogging the clear images, we propose to alleviate the domain gap by considering fog influence and style variation simultaneously. The motivation is based on our finding that the style-related gap and the fog-related gap can be divided and closed respectively, by adding an intermediate domain. Thus, we propose a new pipeline to cumulatively adapt style, fog and the dual-factor (style and fog). Specifically, we devise a unified framework to disentangle the style factor and the fog factor separately, and then the dual-factor from images in different domains. Furthermore, we collaborate the disentanglement of three factors with a novel cumulative loss to thoroughly disentangle these three factors. Our method achieves the state-of-the-art performance on three benchmarks and shows generalization ability in rainy and snowy scenes.

Densely annotating LiDAR point clouds remains too expensive and time-consuming to keep up with the ever growing volume of data. While current literature focuses on fully-supervised performance, developing efficient methods that take advantage of realistic weak supervision have yet to be explored. In this paper, we propose using scribbles to annotate LiDAR point clouds and release ScribbleKITTI, the first scribble-annotated dataset for LiDAR semantic segmentation. Furthermore, we present a pipeline to reduce the performance gap that arises when using such weak annotations. Our pipeline comprises of three stand-alone contributions that can be combined with any LiDAR semantic segmentation model to achieve up to 95.7% of the fully-supervised performance while using only 8% labeled points. Our scribble annotations and code are available at github.com/ouenal/scribblekitti.

Different self-supervised tasks (SSL) reveal different features from the data. The learned feature representations can exhibit different performance for each downstream task. In this light, this work aims to combine Multiple SSL tasks (Multi-SSL) that generalizes wellfor all downstream tasks. For this study, we investigate binaural sounds and image data. For binaural sounds, we propose three SSL tasks namely, spatial alignment, temporal synchronization offore-ground objects and binaural sounds and temporal gap prediction. We investigate several approaches of Multi-SSL and give insights into the downstream task performance on video retrieval, spatial sound super resolution, and semantic prediction using OmniAudio dataset. Our experiments on binaural sound representations demonstrate that Multi-SSL via incremental learning (IL) of SSL tasks outperforms single SSL task models and fully supervised models in the downstream task performance. As a check of applicability on other modalities, we also formulate our Multi-SSL models for image representation learning and we use the recently proposed SSL tasks, MoCov2 and DenseCL. Here, Multi-SSL surpasses recent methods such as MoCov2, DenseCL and DetCo by 2.06%,3.27% and 1.19% on VOC07 classification and +2.83, +1.56 and +1.61 AP on COCO detection.

Multi-Object Tracking (MOT) is most often approached in the tracking-by-detection paradigm, where object detections are associated through time. The association step naturally leads to discrete optimization problems. As these optimization problems are often NP-hard, they can only be solved exactly for small instances on current hardware. Adiabatic quantum computing (AQC) offers a solution for this, as it has the potential to provide a considerable speedup on a range of NP-hard optimization problems in the near future. However, current MOT formulations are unsuitable for quantum computing due to their scaling properties. In this work, we therefore propose the first MOT formulation designed to be solved with AQC. We employ an Ising model that represents the quantum mechanical system implemented on the AQC. We show that our approach is competitive compared with state-of-the-art optimization-based approaches, even when using of-the-shelf integer programming solvers. Finally, we demonstrate that our MOT problem is already solvable on the current generation of real quantum computers for small examples, and analyze the properties of the measured solutions.

Unsupervised domain adaptation (UDA) aims to adapt a model trained on the source domain (e.g. synthetic data) to the target domain (e.g. real-world data) without requiring further annotations on the target domain. This work focuses on UDA for semantic segmentation as real-world pixel-wise annotations are particularly expensive to acquire. As UDA methods for semantic segmentation are usually GPU memory intensive, most previous methods operate only on downscaled images. We question this design as low-resolution predictions often fail to preserve fine details. The alternative of training with random crops of high-resolution images alleviates this problem but falls short in capturing long-range, domain-robust context information. Therefore, we propose HRDA, a multi-resolution training approach for UDA, that combines the strengths of small high-resolution crops to preserve fine segmentation details and large low-resolution crops to capture long-range context dependencies with a learned scale attention, while maintaining a manageable GPU memory footprint. HRDA enables adapting small objects and preserving fine segmentation details. It significantly improves the state-of-the-art performance by 5.5 mIoU for GTA $$\rightarrow $$ Cityscapes and 4.9 mIoU for Synthia $$\rightarrow $$ Cityscapes, resulting in unprecedented 73.8 and 65.8 mIoU, respectively. The implementation is available at github.com/lhoyer/HRDA .

Predicting multimodal future behavior of traffic participants is essential for robotic vehicles to make safe decisions. Existing works explore to directly predict future trajectories based on latent features or utilize dense goal candidates to identify agent's destinations, where the former strategy converges slowly since all motion modes are derived from the same feature while the latter strategy has efficiency issue since its performance highly relies on the density of goal candidates. In this paper, we propose the Motion TRansformer (MTR) framework that models motion prediction as the joint optimization of global intention localization and local movement refinement. Instead of using goal candidates, MTR incorporates spatial intention priors by adopting a small set of learnable motion query pairs. Each motion query pair takes charge of trajectory prediction and refinement for a specific motion mode, which stabilizes the training process and facilitates better multimodal predictions. Experiments show that MTR achieves state-of-the-art performance on both the marginal and joint motion prediction challenges, ranking 1st on the leaderbaords of Waymo Open Motion Dataset. Code will be available at https://github.com/sshaoshuai/MTR.

This work studies Hyperspectral image (HSI) super-resolution (SR). HSI SR is characterized by high-dimensional data and a limited amount of training examples. This raises challenges for training deep neural networks that are known to be data hungry. This work addresses this issue with two contributions. First, we observe that HSI SR and RGB image SR are correlated and develop a novel multi-tasking network to train them jointly so that the auxiliary task RGB image SR can provide additional supervision and regulate the network training. Second, we extend the network to a semi-supervised setting so that it can learn from datasets containing only low-resolution HSIs. With these contributions, our method is able to learn hyperspectral image super-resolution from heterogeneous datasets and lifts the requirement for having a large amount of high resolution (HR) HSI training samples. Extensive experiments on three standard datasets show that our method outperforms existing methods significantly and underpin the relevance of our contributions. Our code can be found at https://github.com/kli8996/HSISR.git.

We address the problem of semantic nighttime image segmentation and improve the state-of-the-art, by adapting daytime models to nighttime without using nighttime annotations. Moreover, we design a new evaluation framework to address the substantial uncertainty of semantics in nighttime images. Our central contributions are: 1) a curriculum framework to gradually adapt semantic segmentation models from day to night through progressively darker times of day, exploiting cross-time-of-day correspondences between daytime images from a reference map and dark images to guide the label inference in the dark domains; 2) a novel uncertainty-aware annotation and evaluation framework and metric for semantic segmentation, including image regions beyond human recognition capability in the evaluation in a principled fashion; 3) the Dark Zurich dataset, comprising 2416 unlabeled nighttime and 2920 unlabeled twilight images with correspondences to their daytime counterparts plus a set of 201 nighttime images with fine pixel-level annotations created with our protocol, which serves as a first benchmark for our novel evaluation. Experiments show that our map-guided curriculum adaptation significantly outperforms state-of-the-art methods on nighttime sets both for standard metrics and our uncertainty-aware metric. Furthermore, our uncertainty-aware evaluation reveals that selective invalidation of predictions can improve results on data with ambiguous content such as our benchmark and profit safety-oriented applications involving invalid inputs.

With the advent of deep learning, many dense prediction tasks, i.e., tasks that produce pixel-level predictions, have seen significant performance improvements. The typical approach is to learn these tasks in isolation, that is, a separate neural network is trained for each individual task. Yet, recent multi-task learning (MTL) techniques have shown promising results w.r.t. performance, computations and/or memory footprint, by jointly tackling multiple tasks through a learned shared representation. In this survey, we provide a well-rounded view on state-of-the-art deep learning approaches for MTL in computer vision, explicitly emphasizing on dense prediction tasks. Our contributions concern the following. First, we consider MTL from a network architecture point-of-view. We include an extensive overview and discuss the advantages/disadvantages of recent popular MTL models. Second, we examine various optimization methods to tackle the joint learning of multiple tasks. We summarize the qualitative elements of these works and explore their commonalities and differences. Finally, we provide an extensive experimental evaluation across a variety of dense prediction benchmarks to examine the pros and cons of the different methods, including both architectural and optimization based strategies.

Image-to-image translation is to map images from a given style to another given style. While exceptionally successful, current methods assume the availability of training images in both source and target domains, which does not always hold in practice. Inspired by humans' reasoning capability of analogy, we propose analogical image translation (AIT) that exploit the concept of gist, for the first time. Given images of two styles in the source domain: A and A', along with images B of the first style in the target domain, learn a model to translate B to B' in the target domain, such that A:A' :: B:B'. AIT is especially useful for translation scenarios in which training data of one style is hard to obtain but training data of the same two styles in another domain is available. For instance, in the case from normal conditions to extreme, rare conditions, obtaining real training images for the latter case is challenging. However, obtaining synthetic data for both cases is relatively easy. In this work, we aim at adding adverse weather effects, more specifically fog, to images taken in clear weather. To circumvent the challenge of collecting real foggy images, AIT learns the gist of translating synthetic clear-weather to foggy images, followed by adding fog effects onto real clear-weather images, without ever seeing any real foggy image. AIT achieves zero-shot image translation capability, whose effectiveness and benefit are demonstrated by the downstream task of semantic foggy scene understanding.

We study low-rank parameterizations of weight matrices with embedded spectral properties in the Deep Learning context. The low-rank property leads to parameter efficiency and permits taking computational shortcuts when computing mappings. Spectral properties are often subject to constraints in optimization problems, leading to better models and stability of optimization. We start by looking at the compact SVD parameterization of weight matrices and identifying redundancy sources in the parameterization. We further apply the Tensor Train (TT) decomposition to the compact SVD components, and propose a non-redundant differentiable parameterization of fixed TT-rank tensor manifolds, termed the Spectral Tensor Train Parameterization (STTP). We demonstrate the effects of neural network compression in the image classification setting, and both compression and improved training stability in the generative adversarial training setting. Project website: www.obukhov.ai/sttp

Open compound domain adaptation (OCDA) is a domain adaptation setting, where target domain is modeled as a compound of multiple unknown homogeneous domains, which brings the advantage of improved generalization to unseen domains. In this work, we propose a principled meta-learning based approach to OCDA for semantic segmentation, MOCDA, by modeling the unlabeled target domain continuously. Our approach consists of four key steps. First, we cluster target domain into multiple sub-target domains by image styles, extracted in an unsupervised manner. Then, different sub-target domains are split into independent branches, for which batch normalization parameters are learnt to treat them independently. A meta-learner is thereafter deployed to learn to fuse sub-target domain-specific predictions, conditioned upon the style code. Meanwhile, we learn to online update the model by model-agnostic meta-learning (MAML) algorithm, thus to further improve generalization. We validate the benefits of our approach by extensive experiments on synthetic-to-real knowledge transfer benchmark, where we achieve the state-of-the-art performance in both compound and open domains.

Training deep networks for semantic segmentation requires large amounts of labeled training data, which presents a major challenge in practice, as labeling segmentation masks is a highly labor-intensive process. To address this issue, we present a framework for semi-supervised semantic segmentation, which is enhanced by self-supervised monocular depth estimation from unlabeled image sequences. In particular, we propose three key contributions: (1) We transfer knowledge from features learned during self-supervised depth estimation to semantic segmentation, (2) we implement a strong data augmentation by blending images and labels using the geometry of the scene, and (3) we utilize the depth feature diversity as well as the level of difficulty of learning depth in a student-teacher framework to select the most useful samples to be annotated for semantic segmentation. We validate the proposed model on the Cityscapes dataset, where all three modules demonstrate significant performance gains, and we achieve state-of-the-art results for semi-supervised semantic segmentation. The implementation is available at https://github.com/lhoyer/improving_segmentation_with_selfsupervised_depth.

In this paper, we address the problem of rain streaks and rain accumulation removal in video, by developing a self-alignment network with transmission-depth consistency. Existing video based deraining methods focus only on rain streak removal, and commonly use optical flow to align the rain video frames. However, besides rain streaks, rain accummulation can considerably degrade visibility; and, optical flow estimation in a rain video is still erroneous, making the deraining performance tend to be inaccurate. Our method employs deformable convolution layers in our encoder to achieve feature-level frame alignment, and hence avoids using optical flow. For rain streaks, our method predicts the current frame from its adjacent frames, such that rain streaks that appear randomly in the temporal domain can be removed. For rain accumulation, our method employs a transmission-depth consistency loss to resolve the ambiguity between the depth and water-droplet density. Our network estimates the depth from consecutive rain-accumulation-removal outputs, and calculates the transmission map using a commonly used physics model. To ensure photometric-temporal and depth-temporal consistencies, our method estimates the camera poses, so that it can warp one frame to its adjacent frames. Experimental results show that our method is effective in removing both rain streaks and rain accumulation, outperforming those of state-of-the-art methods quantitatively and qualitatively.

Domain adaptation for semantic segmentation aims to improve the model performance in the presence of a distribution shift between source and target domain. Leveraging the supervision from auxiliary tasks (such as depth estimation) has the potential to heal this shift because many visual tasks are closely related to each other. However, such a supervision is not always available. In this work, we leverage the guidance from self-supervised depth estimation, which is available on both domains, to bridge the domain gap. On the one hand, we propose to explicitly learn the task feature correlation to strengthen the target semantic predictions with the help of target depth estimation. On the other hand, we use the depth prediction discrepancy from source and target depth decoders to approximate the pixel-wise adaptation difficulty. The adaptation difficulty, inferred from depth, is then used to refine the target semantic segmentation pseudo-labels. The proposed method can be easily implemented into existing segmentation frameworks. We demonstrate the effectiveness of our approach on the benchmark tasks SYNTHIA-to-Cityscapes and GTA-to-Cityscapes, on which we achieve the new state-of-the-art performance of 55.0% and 56.6%, respectively. Our code is available at https://qin.ee/corda.

One challenge of object recognition is to generalize to new domains, to more classes and/or to new modalities. This necessitates methods to combine and reuse existing datasets that may belong to different domains, have partial annotations, and/or have different data modalities. This paper formulates this as a multi-source domain adaptation and label unification problem, and proposes a novel method for it. Our method consists of a partially-supervised adaptation stage and a fully-supervised adaptation stage. In the former, partial knowledge is transferred from multiple source domains to the target domain and fused therein. Negative transfer between unmatching label spaces is mitigated via three new modules: domain attention, uncertainty maximization and attention-guided adversarial alignment. In the latter, knowledge is transferred in the unified label space after a label completion process with pseudolabels. Extensive experiments on three different tasks - image classification, 2D semantic image segmentation, and joint 2D-3D semantic segmentation - show that our method outperforms all competing methods significantly.

This work addresses the challenging task of LiDAR-based 3D object detection in foggy weather. Collecting and annotating data in such a scenario is very time, labor and cost intensive. In this paper, we tackle this problem by simulating physically accurate fog into clear-weather scenes, so that the abundant existing real datasets captured in clear weather can be repurposed for our task. Our contributions are twofold: 1) We develop a physically valid fog simulation method that is applicable to any LiDAR dataset. This unleashes the acquisition of large-scale foggy training data at no extra cost. These partially synthetic data can be used to improve the robustness of several perception methods, such as 3D object detection and tracking or simultaneous localization and mapping, on real foggy data. 2) Through extensive experiments with several state-of-the-art detection approaches, we show that our fog simulation can be leveraged to significantly improve the performance for 3D object detection in the presence of fog. Thus, we are the first to provide strong 3D object detection baselines on the Seeing Through Fog dataset. Our code is available at www.trace.ethz.ch/lidar fog simulation.

Level 5 autonomy for self-driving cars requires a robust visual perception system that can parse input images under any visual condition. However, existing semantic segmentation datasets are either dominated by images captured under normal conditions or are small in scale. To address this, we introduce ACDC, the Adverse Conditions Dataset with Correspondences for training and testing semantic segmentation methods on adverse visual conditions. ACDC consists of a large set of 4006 images which are equally distributed between four common adverse conditions: fog, nighttime, rain, and snow. Each adverse-condition image comes with a high-quality fine pixel-level semantic annotation, a corresponding image of the same scene taken under normal conditions, and a binary mask that distinguishes between intra-image regions of clear and uncertain semantic content. Thus, ACDC supports both standard semantic segmentation and the newly introduced uncertainty-aware semantic segmentation. A detailed empirical study demonstrates the challenges that the adverse domains of ACDC pose to state-of-the-art supervised and unsupervised approaches and indicates the value of our dataset in steering future progress in the field. Our dataset and benchmark are publicly available.

We introduce Task Switching Networks (TSNs), a task-conditioned architecture with a single unified encoder/decoder for efficient multi-task learning. Multiple tasks are performed by switching between them, performing one task at a time. TSNs have a constant number of parameters irrespective of the number of tasks. This scalable yet conceptually simple approach circumvents the overhead and intricacy of task-specific network components in existing works. In fact, we demonstrate for the first time that multi-tasking can be performed with a single task-conditioned decoder. We achieve this by learning task-specific conditioning parameters through a jointly trained task embedding network, encouraging constructive interaction between tasks. Experiments validate the effectiveness of our approach, achieving state-of-the-art results on two challenging multi-task benchmarks, PASCAL-Context and NYUD. Our analysis of the learned task embeddings further indicates a connection to task relationships studied in the recent literature.

End-to-end approaches to autonomous driving commonly rely on expert demonstrations. Although humans are good drivers, they are not good coaches for end-to-end algorithms that demand dense on-policy supervision. On the contrary, automated experts that leverage privileged information can efficiently generate large scale on-policy and off-policy demonstrations. However, existing automated experts for urban driving make heavy use of hand-crafted rules and perform suboptimally even on driving simulators, where ground-truth information is available. To ad-dress these issues, we train a reinforcement learning expert that maps bird’s-eye view images to continuous low-level actions. While setting a new performance upper-bound on CARLA, our expert is also a better coach that provides in-formative supervision signals for imitation learning agents to learn from. Supervised by our reinforcement learning coach, a baseline end-to-end agent with monocular camera-input achieves expert-level performance. Our end-to-end agent achieves a 78% success rate while generalizing to a new town and new weather on the NoCrash-dense bench-mark and state-of-the-art performance on the more challenging CARLA LeaderBoard.

Point cloud semantic segmentation plays an essential role in autonomous driving, providing vital information about drivable surfaces and nearby objects that can aid higher level tasks such as path planning and collision avoidance. While current 3D semantic segmentation networks focus on convolutional architectures that perform great for well represented classes, they show a significant drop in performance for underrepresented classes that share similar geometric features. We propose a novel Detection Aware 3D Semantic Segmentation (DASS) framework that explicitly leverages localization features from an auxiliary 3D object detection task. By utilizing multitask training, the shared feature representation of the network is guided to be aware of per class detection features that aid tackling the differentiation of geometrically similar classes. We additionally provide a pipeline that uses DASS to generate high recall proposals for existing 2-stage detectors and demonstrate that the added supervisory signal can be used to improve 3D orientation estimation capabilities. Extensive experiments on both the SemanticKITTI and KITTI object datasets show that DASS can improve 3D semantic segmentation results of geometrically similar classes up to 37.8% IoU in image FOV while maintaining high precision bird's-eye view (BEV) detection results.

Abstract The role of robots in society keeps expanding, bringing with it the necessity of interacting and communicating with humans. In order to keep such interaction intuitive, we provide automatic wayfinding based on verbal navigational instructions. Our first contribution is the creation of a large-scale dataset with verbal navigation instructions. To this end, we have developed an interactive visual navigation environment based on Google Street View; we further design an annotation method to highlight mined anchor landmarks and local directions between them in order to help annotators formulate typical, human references to those. The annotation task was crowdsourced on the AMT platform, to construct a new Talk2Nav dataset with 10, 714 routes. Our second contribution is a new learning method. Inspired by spatial cognition research on the mental conceptualization of navigational instructions, we introduce a soft dual attention mechanism defined over the segmented language instructions to jointly extract two partial instructions—one for matching the next upcoming visual landmark and the other for matching the local directions to the next landmark. On the similar lines, we also introduce spatial memory scheme to encode the local directional transitions. Our work takes advantage of the advance in two lines of research: mental formalization of verbal navigational instructions and training neural network agents for automatic way finding. Extensive experiments show that our method significantly outperforms previous navigation methods. For demo video, dataset and code, please refer to our project page .

We introduce T-Basis, a novel concept for a compact representation of a set of tensors, each of an arbitrary shape, which is often seen in Neural Networks. Each of the tensors in the set is modeled using Tensor Rings, though the concept applies to other Tensor Networks. Owing its name to the T-shape of nodes in diagram notation of Tensor Rings, T-Basis is simply a list of equally shaped three-dimensional tensors, used to represent Tensor Ring nodes. Such representation allows us to parameterize the tensor set with a small number of parameters (coefficients of the T-Basis tensors), scaling logarithmically with each tensor’s size in the set and linearly with the dimensionality of T-Basis. We evaluate the proposed approach on the task of neural network compression and demonstrate that it reaches high compression rates at acceptable performance drops. Finally, we analyze memory and operation requirements of the compressed networks and conclude that T-Basis networks are equally well suited for training and inference in resource-constrained environments and usage on the edge devices. Project website: obukhov.ai/tbasis.

This paper investigates how end-to-end driving models can be improved to drive more accurately and human-like. To tackle the first issue we exploit semantic and visual maps from HERE Technologies and augment the existing Drive360 dataset with such. The maps are used in an attention mechanism that promotes segmentation confidence masks, thus focusing the network on semantic classes in the image that are important for the current driving situation. Human-like driving is achieved using adversarial learning, by not only minimizing the imitation loss with respect to the human driver but by further defining a discriminator, that forces the driving model to produce action sequences that are human-like. Our models are trained and evaluated on the Drive360 + HERE dataset, which features 60 hours and 3000 km of real-world driving data. Extensive experiments show that our driving models are more accurate and behave more human-like than previous methods.

In this paper, we explore the possibility of achieving a more accurate depth estimation by fusing monocular images and Radar points using a deep neural network. We give a comprehensive study of the fusion between RGB images and Radar measurements from different aspects and proposed a working solution based on the observations. We find that the noise existing in Radar measurements is one of the main key reasons that prevents one from applying the existing fusion methods developed for LiDAR data and images to the new fusion problem between Radar data and images. The experiments are conducted on the nuScenes dataset, which is one of the first datasets which features Camera, Radar, and LiDAR recordings in diverse scenes and weather conditions. Extensive experiments demonstrate that our method outperforms existing fusion methods. We also provide detailed ablation studies to show the effectiveness of each component in our method.

This work studies the problem of predicting the sequence of future actions for surrounding vehicles in real-world driving scenarios. To this aim, we make three main contributions. The first contribution is an automatic method to convert the trajectories recorded in real-world driving scenarios to action sequences with the help of HD maps. The method enables automatic dataset creation for this task from large-scale driving data. Our second contribution lies in applying the method to the well-known traffic agent tracking and prediction dataset Argoverse, resulting in 228,000 action sequences. Additionally, 2,245 action sequences were manually annotated for testing. The third contribution is to propose a novel action sequence prediction method by integrating past positions and velocities of the traffic agents, map information and social context into a single end-to-end trainable neural network. Our experiments prove the merit of the data creation method and the value of the created dataset - prediction performance improves consistently with the size of the dataset and shows that our action prediction method outperforms comparing models.

Most progress in semantic segmentation reports on daytime images taken under favorable illumination conditions. We instead address the problem of semantic segmentation of nighttime images and improve the state-of-the-art, by adapting daytime models to nighttime without using nighttime annotations. Moreover, we design a new evaluation framework to address the substantial uncertainty of semantics in nighttime images. Our central contributions are: 1) a curriculum framework to gradually adapt semantic segmentation models from day to night via labeled synthetic images and unlabeled real images, both for progressively darker times of day, which exploits cross-time-of-day correspondences for the real images to guide the inference of their labels; 2) a novel uncertainty-aware annotation and evaluation framework and metric for semantic segmentation, designed for adverse conditions and including image regions beyond human recognition capability in the evaluation in a principled fashion; 3) the Dark Zurich dataset, which comprises 2416 unlabeled nighttime and 2920 unlabeled twilight images with correspondences to their daytime counterparts plus a set of 151 nighttime images with fine pixel-level annotations created with our protocol, which serves as a first benchmark to perform our novel evaluation. Experiments show that our guided curriculum adaptation significantly outperforms state-of-the-art methods on real nighttime sets both for standard metrics and our uncertainty-aware metric. Furthermore, our uncertainty-aware evaluation reveals that selective invalidation of predictions can lead to better results on data with ambiguous content such as our nighttime benchmark and profit safety-oriented applications which involve invalid inputs.

Considerable progress has been made in semantic scene understanding of road scenes with monocular cameras. It is, however, mainly focused on certain specific classes such as cars, bicyclists and pedestrians. This work investigates traffic cones, an object category crucial for traffic control in the context of autonomous vehicles. 3D object detection using images from a monocular camera is intrinsically an ill-posed problem. In this work, we exploit the unique structure of traffic cones and propose a pipelined approach to solve this problem. Specifically, we first detect cones in images by a modified 2D object detector. Following which the keypoints on a traffic cone are recognized with the help of our deep structural regression network, here, the fact that the cross-ratio is projection invariant is leveraged for network regularization. Finally, the 3D position of cones is recovered via the classical Perspective n-Point algorithm using correspondences obtained from the keypoint regression. Extensive experiments show that our approach can accurately detect traffic cones and estimate their position in the 3D world in real time. The proposed method is also deployed on a real-time, autonomous system. It runs efficiently on the low-power Jetson TX2, providing accurate 3D position estimates, allowing a race-car to map and drive autonomously on an unseen track indicated by traffic cones. With the help of robust and accurate perception, our race-car won both Formula Student Competitions held in Italy and Germany in 2018, cruising at a top speed of 54 km/h on our driverless platform “gotthard driverless” https://youtu. be/HegmIXASKow?t==11694. Visualization of the complete pipeline, mapping and navigation can be found on our project page http://people.ee.ethz.ch/=tracczucrich/Trnfflcf.lonc/.

Exemplar-based dynamic texture synthesis (EDTS) is targeted to generate new samples of high quality that are perceptually similar to a given input dynamic texture exemplar. This paper addresses the issue of learning the synthesizability of dynamic texture samples. Given a dynamic texture sample, how is its possibility of being synthesized by EDTS methods estimated, and what is the most suitable EDTS algorithm to complete the task? To this end, we propose associating dynamic texture samples with synthesizability scores by learning regression models on a compiled dynamic texture dataset annotated in terms of synthesizability. More precisely, we first define the synthesizability of DT samples and characterize them by a set of spatiotemporal features. We then train regression models on the annotated dataset with feature representation to predict the synthesizability scores of the DT samples and learn classifiers to select the most suitable EDTS algorithm. We further complete the selection, partition and synthesizability prediction of the DT samples in a hierarchical scheme. The learned synthesizability is finally applied to detecting synthesizable regions in videos. Both quantitative and qualitative experiments demonstrate that our method can efficiently learn and predict the synthesizability of DT samples.

Object detection typically assumes that training and test data are drawn from an identical distribution, which, however, does not always hold in practice. Such a distribution mismatch will lead to a significant performance drop. In this work, we aim to improve the cross-domain robustness of object detection. We tackle the domain shift on two levels: (1) the image-level shift, such as image style, illumination, etc., and (2) the instance-level shift, such as object appearance, size, etc. We build our approach based on the recent state-of-the-art Faster R-CNN model, and design two domain adaptation components, on image level and instance level, to reduce the domain discrepancy. The two domain adaptation components are based onH-divergence theory, and are implemented by learning a domain classifier in adversarial training manner. The domain classifiers on different levels are further reinforced with a consistency regularization to learn a domain-invariant region proposal network (RPN) in the Faster R-CNN model. We evaluate our newly proposed approach using multiple datasets including Cityscapes, KITTI, SIM10K, etc. The results demonstrate the effectiveness of our proposed approach for robust object detection in various domain shift scenarios.

We investigate the problem of object referring (OR) i.e. to localize a target object in a visual scene coming with a language description. Humans perceive the world more as continued video snippets than as static images, and describe objects not only by their appearance, but also by their spatio-temporal context and motion features. Humans also gaze at the object when they issue a referring expression. Existing works for OR mostly focus on static images only, which fall short in providing many such cues. This paper addresses OR in videos with language and human gaze. To that end, we present a new video dataset for OR, with 30, 000 objects over 5, 000 stereo video sequences annotated for their descriptions and gaze. We further propose a novel network model for OR in videos, by integrating appearance, motion, gaze, and spatio-temporal context into one network. Experimental results show that our method effectively utilizes motion cues, human gaze, and spatio-temporal context. Our method outperforms previous OR methods. For dataset and code, please refer https://people.ee.ethz.ch/~arunv/ORGaze.html.

The example-based dynamic texture synthesis (EDTS) methods have emerged in multitude, dedicated to generating new dynamic textures (DTs) of high quality from an input exemplar. The problem of EDTS has been studied for several decades, but none of the existing synthesis methods are able to tackle all kinds of dynamic textures equally well. Rather than focus on new synthesis methods, we turn to another way to help EDTS by investigating dynamic texture synthesizability - how synthesizable a specific dynamic texture sample is by EDTS. We propose to predict synthesizability score of a given dynamic texture sample, and suggest which EDTS method is best suited to synthesize it. To this end, we compiled a dynamic texture dataset and annotated each DT in terms of synthesizability. We address the problem of learning dynamic texture synthesizability by using regression model to train a predictor on the data collection. More precisely, we first characterize DT samples by a set of spatiotemporal features. Then, based on dynamic texture descriptors, we train regression models to estimate synthesizability scores and use an additional classifier to choose the optimal EDTS methods. The experiments demonstrate that our method can predict the synthesizability of DT samples effectively.

The primary focus of autonomous driving research is to improve driving accuracy. While great progress has been made, state-of-the-art algorithms still fail at times. Such failures may have catastrophic consequences. It therefore is im- portant that automated cars foresee problems ahead as early as possible. This is also of paramount importance if the driver will be asked to take over. We conjecture that failures do not occur randomly. For instance, driving models may fail more likely at places with heavy traffic, at complex intersections, and/or under adverse weather/illumination conditions. This work presents a method to learn to predict the occurrence of these failures, i.e., to assess how difficult a scene is to a given driving model and to possibly give the human driver an early headsup. A camera- based driving model is developed and trained over real driving datasets. The discrepancies between the model's predictions and the human 'ground-truth' maneuvers were then recorded, to yield the 'failure' scores. Experimental results show that the failure score can indeed be learned and predicted. Thus, our prediction method is able to improve the overall safety of an automated driving model by alerting the human driver timely, leading to better human-vehicle collaborative driving.

Object referring has important applications, especially for human-machine interaction. While having received great attention, the task is mainly attacked with written language (text) as input rather than spoken language (speech), which is more natural. This paper investigates Object Referring with Spoken Language (ORSpoken) by presenting two datasets and one novel approach. Objects are annotated with their locations in images, text descriptions and speech descriptions. This makes the datasets ideal for multi-modality learning. The approach is developed by carefully taking down ORSpoken problem into three sub-problems and introducing taskspecific vision-language interactions at the corresponding levels. Experiments show that our method outperforms competing methods consistently and significantly. The approach is also evaluated in the presence of audio noise, showing the efficacy of the proposed vision-language interaction methods in counteracting background noise.

Domain adaptation between diverse source and target domains is challenging, especially in the real-world visual recognition tasks where the images and videos consist of significant variations in viewpoints, illuminations, qualities, etc. In this paper, we propose a new approach for domain generalization and domain adaptation based on exemplar SVMs. Specifically, we decompose the source domain into many subdomains, each of which contains only one positive training sample and all negative samples. Each subdomain is relatively less diverse, and is expected to have a simpler distribution. By training one exemplar SVM for each subdomain, we obtain a set of exemplar SVMs. To further exploit the inherent structure of source domain, we introduce a nuclear-norm based regularizer into the objective function in order to enforce the exemplar SVMs to produce a low-rank output on training samples. In the prediction process, the confident exemplar SVM classifiers are selected and reweigted according to the distribution mismatch between each subdomain and the test sample in the target domain. We formulate our approach based on the logistic regression and least square SVM algorithms, which are referred to as low rank exemplar SVMs (LRE-SVMs) and low rank exemplar least square SVMs (LRE-LSSVMs), respectively. A fast algorithm is also developed for accelerating the training of LRE-LSSVMs. We further extend Domain Adaptation Machine (DAM) to learn an optimal target classifier for domain adaptation, and show that our approach can also be applied to domain adaptation with evolving target domain, where the target data distribution is gradually changing. The comprehensive experiments for object recognition and action recognition demonstrate the effectiveness of our approach for domain generalization and domain adaptation with fixed and evolving target domains.

Progress in Multiple Object Tracking (MOT) has been historically limited by the size of the available datasets. We present an efficient framework to annotate trajectories and use it to produce a MOT dataset of unprecedented size. In our novel path supervision the annotator loosely follows the object with the cursor while watching the video, providing a path annotation for each object in the sequence. Our approach is able to turn such weak annotations into dense box trajectories. Our experiments on existing datasets prove that our framework produces more accurate annotations than the state of the art, in a fraction of the time. We further validate our approach by crowdsourcing the PathTrack dataset, with more than 15,000 person trajectories in 720 sequences. Tracking approaches can benefit training on such large-scale datasets, as did object recognition. We prove this by re-training an off-the-shelf person matching network, originally trained on the MOT15 dataset, almost halving the misclassification rate. Additionally, training on our data consistently improves tracking results, both on our dataset and on MOT15. On the latter, we improve the top-performing tracker (NOMT) dropping the number of ID Switches by 18% and fragments by 5%.

Image segmentation is a key component in many computer vision systems, and it is recovering a prominent spot in the literature as methods improve and overcome their limitations. The outputs of most recent algorithms are in the form of a hierarchical segmentation, which provides segmentation at different scales in a single tree-like structure. Commonly, these hierarchical methods start from some low-level features, and are not aware of the scale information of the different regions in them. As such, one might need to work on many different levels of the hierarchy to find the objects in the scene. This work tries to modify the existing hierarchical algorithm by improving their alignment, that is, by trying to modify the depth of the regions in the tree to better couple depth and scale. To do so, we first train a regressor to predict the scale of regions using mid-level features. We then define the anchor slice as the set of regions that better balance between over-segmentation and under-segmentation. The output of our method is an improved hierarchy, re-aligned by the anchor slice. To demonstrate the power of our method, we perform comprehensive experiments, which show that our method, as a post-processing step, can significantly improve the quality of the hierarchical segmentation representations, and ease the usage of hierarchical image segmentation to high-level vision tasks such as object segmentation. We also prove that the improvement generalizes well across different algorithms and datasets, with a low computational cost.

The SVM+ approach has shown excellent performance in visual recognition tasks for exploiting privileged information in the training data. In this paper, we propose two efficient algorithms for solving the linear and kernel SVM+, respectively. For linear SVM+, we absorb the bias term into the weight vector, and formulate a new optimization problem with simpler constraints in the dual form. Then, we develop an efficient dual coordinate descent algorithm to solve the new optimization problem. For kernel SVM+, we further apply the l2-loss, which leads to a simpler optimization problem in the dual form with only half of dual variables when compared with the dual form of the original SVM+ method. More interestingly, we show that our new dual problem can be efficiently solved by using the SMO algorithm of the one-class SVM problem. Comprehensive experiments on three datasets clearly demonstrate that our proposed algorithms achieve significant speed-up than the state-of-the-art solvers for linear and kernel SVM+.

Despite the great advances made in the field of image super-resolution (ISR) during the last years, the performance has merely been evaluated perceptually. Thus, it is still unclear whether ISR is helpful for other vision tasks. In this paper, we present the first comprehensive study and analysis of the usefulness of ISR for other vision applications. In particular, six ISR methods are evaluated on four popular vision tasks, namely edge detection, semantic image segmentation, digit recognition, and scene recognition. We show that applying ISR to input images of other vision systems does improve their performance when the input images are of low-resolution. We also study the correlation between four standard perceptual evaluation criteria (namely PSNR, SSIM, IFC, and NQM) and the usefulness of ISR to the vision tasks. Experiments show that they correlate well with each other in general, but perceptual criteria are still not accurate enough to be used as full proxies for the usefulness. We hope this work will inspire the community to evaluate ISR methods also in real vision applications, and to adopt ISR as a pre-processing step of other vision tasks if the resolution of their input images is low.

Multi-target tracking (MTT) is the task of localizing objects of interest in a video and associating them through time. Accurate affinity measures between object detections is crucial for MTT. Previous methods use simple affinity measures, based on heuristics, that are unable to track through occlusions and missing detections. To address this problem, this paper proposes a novel affinity measure by leveraging the power of single-target visual tracking (VT), which has proven reliable to locally track objects of interest given a bounding-box initialization. In particular, given two detections at different frames, we perform VT starting from each of them and towards the frame of the other. We then learn a metric with features extracted from the behaviours (e.g. overlaps and distances) of the two tracking trajectories. By plugging our learned affinity into the standard MTT framework, we are able to cope with occlusions and large amounts of missing or inaccurate detections. We evaluate our method on public datasets, including the popular MOT benchmark, and show improvements over previously published methods.

Metric learning has proved very successful. However, human annotations are necessary. In this paper, we propose an unsupervised method, dubbed Metric Imitation (MI), where metrics over cheap features (target features, TFs) are learned by imitating the standard metrics over more sophisticated, off-the-shelf features (source features, SFs) by transferring view-independent property manifold structures. In particular, MI consists of: 1) quantifying the properties of source metrics as manifold geometry, 2) transferring the manifold from source domain to target domain, and 3) learning a mapping of TFs so that the manifold is approximated as well as possible in the mapped feature domain. MI is useful in at least two scenarios where: 1) TFs are more efficient computationally and in terms of memory than SFs; and 2) SFs contain privileged information, but are not available during testing. For the former, MI is evaluated on image clustering, category-based image retrieval, and instance-based object retrieval, with three SFs and three TFs. For the latter, MI is tested on the task of example-based image super-resolution, where high-resolution patches are taken as SFs and low-resolution patches as TFs. Experiments show that MI is able to provide good metrics while avoiding expensive data labeling efforts and that it achieves state-of-the-art performance for image super-resolution. In addition, manifold transfer is an interesting direction of transfer learning.

We present a novel vanishing point (VP) detection and tracking algorithm for calibrated monocular image sequences. Previous VP detection and tracking methods usually assume known camera poses for all frames or detect and track separately. We advance the state-of-the-art by combining VP extraction on a Gaussian sphere with recent advances in multi-target tracking on probabilistic occupancy fields. The solution is obtained by solving a Linear Program (LP). This enables the joint detection and tracking of multiple VPs over sequences. Unlike existing works we do not need known camera poses, and at the same time avoid detecting and tracking in separate steps. We also propose an extension to enforce VP orthogonality. We augment an existing video dataset consisting of 48 monocular videos with multiple annotated VPs in 14448 frames for evaluation. Although the method is designed for unknown camera poses, it is also helpful in scenarios with known poses, since a multi-frame approach in VP detection helps to regularize in frames with weak VP line support.

Abstract Learning regressors from low‐resolution patches to high‐resolution patches has shown promising results for image super‐resolution. We observe that some regressors are better at dealing with certain cases, and others with different cases. In this paper, we jointly learn a collection of regressors, which collectively yield the smallest super‐resolving error for all training data. After training, each training sample is associated with a label to indicate its ‘best’ regressor, the one yielding the smallest error. During testing, our method bases on the concept of ‘adaptive selection’ to select the most appropriate regressor for each input patch. We assume that similar patches can be super‐resolved by the same regressor and use a fast, approximate kNN approach to transfer the labels of training patches to test patches. The method is conceptually simple and computationally efficient, yet very effective. Experiments on four datasets show that our method outperforms competing methods.

Example-based texture synthesis (ETS) has been widely used to generate high quality textures of desired sizes from a small example. However, not all textures are equally well reproducible that way. We predict how synthesizable a particular texture is by ETS. We introduce a dataset (21, 302 textures) of which all images have been annotated in terms of their synthesizability. We design a set of texture features, such as 'textureness', homogeneity, repetitiveness, and irregularity, and train a predictor using these features on the data collection. This work is the first attempt to quantify this image property, and we find that texture synthesizability can be learned and predicted. We use this insight to trim images to parts that are more synthesizable. Also we suggest which texture synthesis method is best suited to synthesise a given texture. Our approach can be seen as 'winner-uses-all': picking one method among several alternatives, ending up with an overall superior ETS method. Such strategy could also be considered for other vision tasks: rather than building an even stronger method, choose from existing methods based on some simple preprocessing.

Dictionary learning (DL) for sparse coding has shown promising results in classification tasks, while how to adaptively build the relationship between dictionary atoms and class labels is still an important open question. The existing dictionary learning approaches simply fix a dictionary atom to be either class-specific or shared by all classes beforehand, ignoring that the relationship needs to be updated during DL. To address this issue, in this paper we propose a novel latent dictionary learning (LDL) method to learn a discriminative dictionary and build its relationship to class labels adaptively. Each dictionary atom is jointly learned with a latent vector, which associates this atom to the representation of different classes. More specifically, we introduce a latent representation model, in which discrimination of the learned dictionary is exploited via minimizing the within-class scatter of coding coefficients and the latent-value weighted dictionary coherence. The optimal solution is efficiently obtained by the proposed solving algorithm. Correspondingly, a latent sparse representation based classifier is also presented. Experimental results demonstrate that our algorithm outperforms many recently proposed sparse representation and dictionary learning approaches for action, gender and face recognition.

This paper investigates the problem of semi-supervised classification. Unlike previous methods to regularize classifying boundaries with unlabeled data, our method learns a new image representation from all available data (labeled and unlabeled) and performs plain supervised learning with the new feature. In particular, an ensemble of image prototype sets are sampled automatically from the available data, to represent a rich set of visual categories/attributes. Discriminative functions are then learned on these prototype sets, and image are represented by the concatenation of their projected values onto the prototypes (similarities to them) for further classification. Experiments on four standard datasets show three interesting phenomena: (1) our method consistently outperforms previous methods for semi-supervised image classification, (2) our method lets itself combine well with these methods, and (3) our method works well for self-taught image classification where unlabeled data are not coming from the same distribution as labeled ones, but rather from a random collection of images.

There is an increased interest in the efficient creation of city models, be it virtual or as-built. We present a method for synthesizing complex, photo-realistic facade images, from a single example. After parsing the example image into its semantic components, a tiling for it is generated. Novel tilings can then be created, yielding facade textures with different dimensions or with occluded parts in painted. A genetic algorithm guides the novel facades as well as in painted parts to be consistent with the example, both in terms of their overall structure and their detailed textures. Promising results for multiple standard datasets - in particular for the different building styles they contain - demonstrate the potential of the method.

We introduce the hierarchical Markov aspect model (HMAM), a computationally efficient graphical model for densely labeling large remote sensing images with their underlying terrain classes. HMAM resolves local ambiguities efficiently by combining the benefits of quadtree representations and aspect models-the former incorporate multiscale visual features and hierarchical smoothing to provide improved local label consistency, while the latter sharpen the labelings by focusing them on the classes that are most relevant for the broader local image context. The full HMAM model takes a grid of local hierarchical Markov quadtrees over image patches and augments it by incorporating a probabilistic latent semantic analysis aspect model over a larger local image tile at each level of the quadtree forest. Bag-of-word visual features are extracted for each level and patch, and given these, the parent-child transition probabilities from the quadtree and the label probabilities from the tile-level aspect models, an efficient forwards-backwards inference pass allows local posteriors for the class labels to be obtained for each patch. Variational expectation-maximization is then used to train the complete model from either pixel-level or tile-keyword-level labelings. Experiments on a complete TerraSAR-X synthetic aperture radar terrain map with pixel-level ground truth show that HMAM is both accurate and efficient, providing significantly better results than comparable single-scale aspect models with only a modest increase in training and test complexity. Keyword-level training greatly reduces the cost of providing training data with little loss of accuracy relative to pixel-level training.

This letter presents a method for satellite image classification aiming at the following two objectives: 1) involving visual attention into the satellite image classification; biologically inspired saliency information is exploited in the phase of the image representation, making our method more concentrated on the interesting objects and structures, and 2) handling the satellite image classification without the learning phase. A two-layer sparse coding (TSC) model is designed to discover the “true” neighbors of the images and bypass the intensive learning phase of the satellite image classification. The underlying philosophy of the TSC is that an image can be more sparsely reconstructed via the images (sparse I) belonging to the same category (sparse II). The images are classified according to a newly defined “image-to-category” similarity based on the coding coefficients. Requiring no training phase, our method achieves very promising results. The experimental comparisons are shown on a real satellite image database.

In this letter, we propose a theoretically and computationally simple feature for synthetic aperture radar (SAR) image classification, the multilevel local pattern histogram (MLPH). The MLPH describes the size distributions of bright, dark, and homogenous patterns appearing in a moving window at various contrasts; these patterns are the elementary properties of SAR image texture. The MLPH is a very powerful descriptor of SAR images because it captures both local and global structural information. Additionally, it is robust to speckle noise. Experiments on a TerraSAR-X data set demonstrate that MLPH significantly outperforms four other widely used features in SAR image classification.

This paper presents a method for unsupervised scene categorization. Our method aims at two objectives: (1) automatic feature selection for different scene categories. We represent images in a heterogeneous feature space to account for the large variabilities of different scene categories. Then, we use the information projection strategy to pursue features which are both informative and discriminative, and simultaneously learn a generative model for each category. (2) automatic cluster number selection for the whole image set to be categorized. By treating each image as a vertex in a graph, we formulate unsupervised scene categorization as a graph partition problem under the Bayesian framework. Then, we use a cluster sampling strategy to do the partition (i.e. categorization) in which the cluster number is selected automatically for the globally optimal clustering in terms of maximizing a Bayesian posterior probability. In experiments, we test two datasets, LHI 8 scene categories and MIT 8 scene categories, and obtain state-of-the-art results.

In this paper, we present a fast approach to obtain semantic scene segmentation with high precision. We employ a two-stage classifier to label all image pixels. First, we use the regularized logistic regression to combine different appearance-based features and the improved spatial layout of labeling information. In the second stage, we incorporate the local, regional and global cues into a conditional random field model to provide a final segmentation, and a fast max-margin training method is employed to learn the parameters of the model quickly. The comparison experiments on four multi-class image segmentation databases show that our approach can achieve comparable semantic segmentation results and work faster than that of the state-of-the-art approaches.

In this paper, we propose a three-layer spatial sparse coding (TSSC) for image classification, aiming at three objectives: naturally recognizing image categories without learning phase, naturally involving spatial configurations of images, and naturally counteracting the intra-class variances. The method begins by representing the test images in a spatial pyramid as the to-be-recovered signals, and taking all sampled image patches at multiple scales from the labeled images as the bases. Then, three sets of coefficients are involved into the cardinal sparse coding to get the TSSC, one to penalize spatial inconsistencies of the pyramid cells and the corresponding selected bases, one to guarantee the sparsity of selected images, and the other to guarantee the sparsity of selected categories. Finally, the test images are classified according to a simple image-to-category similarity defined on the coding coefficients. In experiments, we test our method on two publicly available datasets and achieve significantly more accurate results than the conventional sparse coding with only a modest increase in computational complexity.