Self-training based on pseudo-labels has emerged as a dominant approach for addressing conditional distribution shifts in unsupervised domain adaptation (UDA) for semantic segmentation problems. A notable drawback, however, is that this family of approaches is susceptible to erroneous pseudo labels that arise from confirmation biases in the source domain and that manifest as nuisance factors in the target domain. A possible source for this mismatch is the reliance on only photometric cues provided by RGB image inputs, which may ultimately lead to sub-optimal adaptation. To mitigate the effect of mismatched pseudo-labels, we propose to incorporate structural cues from auxiliary modalities, such as depth, to regularise conventional self-training objectives. Specifically, we introduce a contrastive pixel-level objectness constraint that pulls the pixel representations within a region of an object instance closer, while pushing those from different object categories apart. To obtain object regions consistent with the true underlying object, we extract information from both depth maps and RGB-images in the form of multimodal clustering. Crucially, the objectness constraint is agnostic to the ground-truth semantic labels and, hence, appropriate for unsupervised domain adaptation. In this work, we show that our regularizer significantly improves top performing self-training methods (by up to $2$ points) in various UDA benchmarks for semantic segmentation. We include all code in the supplementary.
Moving object segmentation (MOS) in dynamic scenes is an important, challenging, but under-explored research topic for autonomous driving, especially for sequences obtained from moving ego vehicles. Most segmentation methods leverage motion cues obtained from optical flow maps. However, since these methods are often based on optical flows that are pre-computed from successive RGB frames, this neglects the temporal consideration of events occurring within the inter-frame, consequently constraining its ability to discern objects exhibiting relative staticity but genuinely in motion. To address these limitations, we propose to exploit event cameras for better video understanding, which provide rich motion cues without relying on optical flow. To foster research in this area, we first introduce a novel large-scale dataset called DSEC-MOS for moving object segmentation from moving ego vehicles, which is the first of its kind. For benchmarking, we select various mainstream methods and rigorously evaluate them on our dataset. Subsequently, we devise EmoFormer, a novel network able to exploit the event data. For this purpose, we fuse the event temporal prior with spatial semantic maps to distinguish genuinely moving objects from the static background, adding another level of dense supervision around our object of interest. Our proposed network relies only on event data for training but does not require event input during inference, making it directly comparable to frame-only methods in terms of efficiency and more widely usable in many application cases. The exhaustive comparison highlights a significant performance improvement of our method over all other methods. The source code and dataset are publicly available at: https://github.com/ZZY-Zhou/DSEC-MOS.
Despite radar's popularity in the automotive industry, for fusion-based 3D object detection, most existing works focus on LiDAR and camera fusion. In this paper, we propose TransCAR, a Transformer-based Camera-And-Radar fusion solution for 3D object detection. Our TransCAR consists of two modules. The first module learns 2D features from surround-view camera images and then uses a sparse set of 3D object queries to index into these 2D features. The vision-updated queries then interact with each other via transformer self-attention layer. The second module learns radar features from multiple radar scans and then applies transformer decoder to learn the interactions between radar features and vision-updated queries. The cross-attention layer within the transformer decoder can adaptively learn the soft-association between the radar features and vision-updated queries instead of hard-association based on sensor calibration only. Finally, our model estimates a bounding box per query using set-to-set Hungarian loss, which enables the method to avoid non-maximum suppression. TransCAR improves the velocity estimation using the radar scans without temporal information. The superior experimental results of our TransCAR on the challenging nuScenes datasets illustrate that our TransCAR outperforms state-of-the-art Camera-Radar fusion-based 3D object detection approaches.
Unsupervised learning of object-centric representations in dynamic visual scenes is challenging. Unlike most previous approaches that learn to decompose 2D images, we present DynaVol, a 3D scene generative model that unifies geometric structures and object-centric learning in a differentiable volume rendering framework. The key idea is to perform object-centric voxelization to capture the 3D nature of the scene, which infers the probability distribution over objects at individual spatial locations. These voxel features evolve over time through a canonical-space deformation function, forming the basis for global representation learning via slot attention. The voxel features and global features are complementary and are both leveraged by a compositional NeRF decoder for volume rendering. DynaVol remarkably outperforms existing approaches for unsupervised dynamic scene decomposition. Once trained, the explicitly meaningful voxel features enable additional capabilities that 2D scene decomposition methods cannot achieve: it is possible to freely edit the geometric shapes or manipulate the motion trajectories of the objects.
Current multi-modal object detection approaches focus on the vehicle domain and are limited in the perception range and the processing capabilities. Roadside sensor units (RSUs) introduce a new domain for perception systems and leverage altitude to observe traffic. Cameras and LiDARs mounted on gantry bridges increase the perception range and produce a full digital twin of the traffic. In this work, we introduce InfraDet3D, a multi-modal 3D object detector for roadside infrastructure sensors. We fuse two LiDARs using early fusion and further incorporate detections from monocular cameras to increase the robustness and to detect small objects. Our monocular 3D detection module uses HD maps to ground object yaw hypotheses, improving the final perception results. The perception framework is deployed on a real-world intersection that is part of the A9 Test Stretch in Munich, Germany. We perform several ablation studies and experiments and show that fusing two LiDARs with two cameras leads to an improvement of +1.90 mAP compared to a camera-only solution. We evaluate our results on the A9 infrastructure dataset and achieve 68.48 mAP on the test set. The dataset and code will be available at https://a9-dataset.com to allow the research community to further improve the perception results and make autonomous driving safer.
In this paper, we propose an advanced methodology for the detection of 3D objects and precise estimation of their spatial positions from a single image. Unlike conventional frameworks that rely solely on center-point and dimension predictions, our research leverages a deep convolutional neural network-based 3D object weighted orientation regression paradigm. These estimates are then seamlessly integrated with geometric constraints obtained from a 2D bounding box, resulting in derivation of a comprehensive 3D bounding box. Our novel network design encompasses two key outputs. The first output involves the estimation of 3D object orientation through the utilization of a discrete-continuous loss function. Simultaneously, the second output predicts objectivity-based confidence scores with minimal variance. Additionally, we also introduce enhancements to our methodology through the incorporation of lightweight residual feature extractors. By combining the derived estimates with the geometric constraints inherent in the 2D bounding box, our approach significantly improves the accuracy of 3D object pose determination, surpassing baseline methodologies. Our method is rigorously evaluated on the KITTI 3D object detection benchmark, demonstrating superior performance.
Weakly supervised object detection (WSOD) aims at learning precise object detectors with only image-level tags. In spite of intensive research on deep learning (DL) approaches over the past few years, there is still a significant performance gap between WSOD and fully supervised object detection. In fact, most existing WSOD methods only consider the visual appearance of each region proposal but ignore employing the useful context information in the image. To this end, this paper proposes an interactive end-to-end WSDO framework called JLWSOD with two innovations: i) two types of WSOD-specific context information (i.e., instance-wise correlation andsemantic-wise correlation) are proposed and introduced into WSOD framework; ii) an interactive graph contrastive learning (iGCL) mechanism is designed to jointly optimize the visual appearance and context information for better WSOD performance. Specifically, the iGCL mechanism takes full advantage of the complementary interpretations of the WSOD, namely instance-wise detection and semantic-wise prediction tasks, forming a more comprehensive solution. Extensive experiments on the widely used PASCAL VOC and MS COCO benchmarks verify the superiority of JLWSOD over alternative state-of-the-art approaches and baseline models (improvement of 3.6%~23.3% on mAP and 3.4%~19.7% on CorLoc, respectively).
This paper shows that it is possible to learn models for monocular 3D reconstruction of articulated objects (e.g., horses, cows, sheep), using as few as 50-150 images labeled with 2D keypoints. Our proposed approach involves training category-specific keypoint estimators, generating 2D keypoint pseudo-labels on unlabeled web images, and using both the labeled and self-labeled sets to train 3D reconstruction models. It is based on two key insights: (1) 2D keypoint estimation networks trained on as few as 50-150 images of a given object category generalize well and generate reliable pseudo-labels; (2) a data selection mechanism can automatically create a "curated" subset of the unlabeled web images that can be used for training -- we evaluate four data selection methods. Coupling these two insights enables us to train models that effectively utilize web images, resulting in improved 3D reconstruction performance for several articulated object categories beyond the fully-supervised baseline. Our approach can quickly bootstrap a model and requires only a few images labeled with 2D keypoints. This requirement can be easily satisfied for any new object category. To showcase the practicality of our approach for predicting the 3D shape of arbitrary object categories, we annotate 2D keypoints on giraffe and bear images from COCO -- the annotation process takes less than 1 minute per image.
In this paper, we introduce a new sequence-to-sequence learning framework for RGB-based and multi-modal object tracking. First, we present SeqTrack for RGB-based tracking. It casts visual tracking as a sequence generation task, forecasting object bounding boxes in an autoregressive manner. This differs from previous trackers, which depend on the design of intricate head networks, such as classification and regression heads. SeqTrack employs a basic encoder-decoder transformer architecture. The encoder utilizes a bidirectional transformer for feature extraction, while the decoder generates bounding box sequences autoregressively using a causal transformer. The loss function is a plain cross-entropy. Second, we introduce SeqTrackv2, a unified sequence-to-sequence framework for multi-modal tracking tasks. Expanding upon SeqTrack, SeqTrackv2 integrates a unified interface for auxiliary modalities and a set of task-prompt tokens to specify the task. This enables it to manage multi-modal tracking tasks using a unified model and parameter set. This sequence learning paradigm not only simplifies the tracking framework, but also showcases superior performance across 14 challenging benchmarks spanning five single- and multi-modal tracking tasks. The code and models are available at https://github.com/chenxin-dlut/SeqTrackv2.
Human-object interactions (HOIs) are crucial for human-centric scene understanding applications such as human-centric visual generation, AR/VR, and robotics. Since existing methods mainly explore capturing HOIs, rendering HOI remains less investigated. In this paper, we address this challenge in HOI animation from a compositional perspective, i.e., animating novel HOIs including novel interaction, novel human and/or novel object driven by a novel pose sequence. Specifically, we adopt neural human-object deformation to model and render HOI dynamics based on implicit neural representations. To enable the interaction pose transferring among different persons and objects, we then devise a new compositional conditional neural radiance field (or CC-NeRF), which decomposes the interdependence between human and object using latent codes to enable compositionally animation control of novel HOIs. Experiments show that the proposed method can generalize well to various novel HOI animation settings. Our project page is https://zhihou7.github.io/CHONA/
Deep Reinforcement Learning has shown significant progress in extracting useful representations from high-dimensional inputs albeit using hand-crafted auxiliary tasks and pseudo rewards. Automatically learning such representations in an object-centric manner geared towards control and fast adaptation remains an open research problem. In this paper, we introduce a method that tries to discover meaningful features from objects, translating them to temporally coherent "question" functions and leveraging the subsequent learned general value functions for control. We compare our approach with state-of-the-art techniques alongside other ablations and show competitive performance in both stationary and non-stationary settings. Finally, we also investigate the discovered general value functions and through qualitative analysis show that the learned representations are not only interpretable but also, centered around objects that are invariant to changes across tasks facilitating fast adaptation.
This research paper compares two neural-network-based adaptive controllers, namely the Hybrid Deep Learning Neural Network Controller (HDLNNC) and the Adaptive Model Predictive Control with Nonlinear Prediction and Linearization along the Predicted Trajectory (AMPC-NPLPT), for controlling a nonlinear object with delay. Specifically, the study investigates the effect of delay on the accuracy of the two controllers. The experimental results demonstrate that the AMPC-NPLPT approach outperforms HDLNNC regarding control accuracy for the given nonlinear object control problem.
Reconstructing the shape and spatially varying surface appearances of a physical-world object as well as its surrounding illumination based on 2D images (e.g., photographs) of the object has been a long-standing problem in computer vision and graphics. In this paper, we introduce an accurate and highly efficient object reconstruction pipeline combining neural based object reconstruction and physics-based inverse rendering (PBIR). Our pipeline firstly leverages a neural SDF based shape reconstruction to produce high-quality but potentially imperfect object shape. Then, we introduce a neural material and lighting distillation stage to achieve high-quality predictions for material and illumination. In the last stage, initialized by the neural predictions, we perform PBIR to refine the initial results and obtain the final high-quality reconstruction of object shape, material, and illumination. Experimental results demonstrate our pipeline significantly outperforms existing methods quality-wise and performance-wise.
Recently, 3D object detection has attracted significant attention and achieved continuous improvement in real road scenarios. The environmental information is collected from a single sensor or multi-sensor fusion to detect interested objects. However, most of the current 3D object detection approaches focus on developing advanced network architectures to improve the detection precision of the object rather than considering the dynamic driving scenes, where data collected from sensors equipped in the vehicle contain various perturbation features. As a result, existing work cannot still tackle the perturbation issue. In order to solve this problem, we propose a group equivariant bird's eye view network (GeqBevNet) based on the group equivariant theory, which introduces the concept of group equivariant into the BEV fusion object detection network. The group equivariant network is embedded into the fused BEV feature map to facilitate the BEV-level rotational equivariant feature extraction, thus leading to lower average orientation error. In order to demonstrate the effectiveness of the GeqBevNet, the network is verified on the nuScenes validation dataset in which mAOE can be decreased to 0.325. Experimental results demonstrate that GeqBevNet can extract more rotational equivariant features in the 3D object detection of the actual road scene and improve the performance of object orientation prediction.
Depth completion and object detection are two crucial tasks often used for aerial 3D mapping, path planning, and collision avoidance of Uncrewed Aerial Vehicles (UAVs). Common solutions include using measurements from a LiDAR sensor; however, the generated point cloud is often sparse and irregular and limits the system's capabilities in 3D rendering and safety-critical decision-making. To mitigate this challenge, information from other sensors on the UAV (viz., a camera used for object detection) is utilized to help the depth completion process generate denser 3D models. Performing both aerial depth completion and object detection tasks while fusing the data from the two sensors poses a challenge to resource efficiency. We address this challenge by proposing a novel approach to jointly execute the two tasks in a single pass. The proposed method is based on an encoder-focused multi-task learning model that exposes the two tasks to jointly learned features. We demonstrate how semantic expectations of the objects in the scene learned by the object detection pathway can boost the performance of the depth completion pathway while placing the missing depth values. Experimental results show that the proposed multi-task network outperforms its single-task counterpart, particularly when exposed to defective inputs.
Language-based object detection is a promising direction towards building a natural interface to describe objects in images that goes far beyond plain category names. While recent methods show great progress in that direction, proper evaluation is lacking. With OmniLabel, we propose a novel task definition, dataset, and evaluation metric. The task subsumes standard- and open-vocabulary detection as well as referring expressions. With more than 28K unique object descriptions on over 25K images, OmniLabel provides a challenging benchmark with diverse and complex object descriptions in a naturally open-vocabulary setting. Moreover, a key differentiation to existing benchmarks is that our object descriptions can refer to one, multiple or even no object, hence, providing negative examples in free-form text. The proposed evaluation handles the large label space and judges performance via a modified average precision metric, which we validate by evaluating strong language-based baselines. OmniLabel indeed provides a challenging test bed for future research on language-based detection.
Object detection is one of the key tasks in many applications of computer vision. Deep Neural Networks (DNNs) are undoubtedly a well-suited approach for object detection. However, such DNNs need highly adapted hardware together with hardware-specific optimization to guarantee high efficiency during inference. This is especially the case when aiming for efficient object detection in video streaming applications on limited hardware such as edge devices. Comparing vendor-specific hardware and related optimization software pipelines in a fair experimental setup is a challenge. In this paper, we propose a framework that uses a host computer with a host software application together with a light-weight interface based on the Message Queuing Telemetry Transport (MQTT) protocol. Various different target devices with target apps can be connected via MQTT with this host computer. With well-defined and standardized MQTT messages, object detection results can be reported to the host computer, where the results are evaluated without harming or influencing the processing on the device. With this quite generic framework, we can measure the object detection performance, the runtime, and the energy efficiency at the same time. The effectiveness of this framework is demonstrated in multiple experiments that offer deep insights into the optimization of DNNs.
Recently, indiscernible/camouflaged scene understanding has attracted lots of research attention in the vision community. We further advance the frontier of this field by systematically studying a new challenge named indiscernible object counting (IOC), the goal of which is to count objects that are blended with respect to their surroundings. Due to a lack of appropriate IOC datasets, we present a large-scale dataset IOCfish5K which contains a total of 5,637 high-resolution images and 659,024 annotated center points. Our dataset consists of a large number of indiscernible objects (mainly fish) in underwater scenes, making the annotation process all the more challenging. IOCfish5K is superior to existing datasets with indiscernible scenes because of its larger scale, higher image resolutions, more annotations, and denser scenes. All these aspects make it the most challenging dataset for IOC so far, supporting progress in this area. Benefiting from the recent advancements of depth estimation foundation models, we construct high-quality depth maps for IOCfish5K by generating pseudo labels using the Depth Anything V2 model. The RGB-D version of IOCfish5K is named IOCfish5K-D. For benchmarking purposes on IOCfish5K, we select 14 mainstream methods for object counting and carefully evaluate them. For multimodal IOCfish5K-D, we evaluate other 4 popular multimodal counting methods. Furthermore, we propose IOCFormer, a new strong baseline that combines density and regression branches in a unified framework and can effectively tackle object counting under concealed scenes. We also propose IOCFormer-D to enable the effective usage of depth modality in helping detect and count objects hidden in their environments. Experiments show that IOCFormer and IOCFormer-D achieve state-of-the-art scores on IOCfish5K and IOCfish5K-D, respectively.
Transformers have shown promising progress in various visual object detection tasks, including monocular 2D/3D detection and surround-view 3D detection. More importantly, the attention mechanism in the Transformer model and the 3D information extraction in binocular stereo are both similarity-based. However, directly applying existing Transformer-based detectors to binocular stereo 3D object detection leads to slow convergence and significant precision drops. We argue that a key cause of that defect is that existing Transformers ignore the binocular-stereo-specific image correspondence information. In this paper, we explore the model design of Transformers in binocular 3D object detection, focusing particularly on extracting and encoding task-specific image correspondence information. To achieve this goal, we present TS3D, a Transformer-based Stereo-aware 3D object detector. In the TS3D, a Disparity-Aware Positional Encoding (DAPE) module is proposed to embed the image correspondence information into stereo features. The correspondence is encoded as normalized sub-pixel-level disparity and is used in conjunction with sinusoidal 2D positional encoding to provide the 3D location information of the scene. To enrich multi-scale stereo features, we propose a Stereo Preserving Feature Pyramid Network (SPFPN). The SPFPN is designed to preserve the correspondence information while fusing intra-scale and aggregating cross-scale stereo features. Our proposed TS3D achieves a 41.29% Moderate Car detection average precision on the KITTI test set and takes 88 ms to detect objects from each binocular image pair. It is competitive with advanced counterparts in terms of both precision and inference speed.
This paper proposes a Robust and Efficient Memory Network, referred to as REMN, for studying semi-supervised video object segmentation (VOS). Memory-based methods have recently achieved outstanding VOS performance by performing non-local pixel-wise matching between the query and memory. However, these methods have two limitations. 1) Non-local matching could cause distractor objects in the background to be incorrectly segmented. 2) Memory features with high temporal redundancy consume significant computing resources. For limitation 1, we introduce a local attention mechanism that tackles the background distraction by enhancing the features of foreground objects with the previous mask. For limitation 2, we first adaptively decide whether to update the memory features depending on the variation of foreground objects to reduce temporal redundancy. Second, we employ a dynamic memory bank, which uses a lightweight and differentiable soft modulation gate to decide how many memory features need to be removed in the temporal dimension. Experiments demonstrate that our REMN achieves state-of-the-art results on DAVIS 2017, with a $\mathcal{J\&F}$ score of 86.3% and on YouTube-VOS 2018, with a $\mathcal{G}$ over mean of 85.5%. Furthermore, our network shows a high inference speed of 25+ FPS and uses relatively few computing resources.