We investigate the effect of style and category similarity in multiple datasets used for object detection pretraining. We find that including an additional stage of object-detection pretraining can increase the detection performance considerably. While our experiments suggest that style similarities between pre-training and target datasets are less important than matching categories, further experiments are needed to verify this hypothesis.
The Odeuropa Challenge on Olfactory Object Recognition aims to foster the development of object detection in the visual arts and to promote an olfactory perspective on digital heritage. Object detection in historical artworks is particularly challenging due to varying styles and artistic periods. Moreover, the task is complicated due to the particularity and historical variance of predefined target objects, which exhibit a large intra-class variance, and the long tail distribution of the dataset labels, with some objects having only very few training examples. These challenges should encourage participants to create innovative approaches using domain adaptation or few-shot learning. We provide a dataset of 2647 artworks annotated with 20 120 tightly fit bounding boxes that are split into a training and validation set (public). A test set containing 1140 artworks and 15 480 annotations is kept private for the challenge evaluation.
Many real-world tasks, from house-cleaning to cooking, can be formulated as multi-object rearrangement problems -- where an agent needs to get specific objects into appropriate goal states. For such problems, we focus on the setting that assumes a pre-specified goal state, availability of perfect manipulation and object recognition capabilities, and a static map of the environment but unknown initial location of objects to be rearranged. Our goal is to enable home-assistive intelligent agents to efficiently plan for rearrangement under such partial observability. This requires efficient trade-offs between exploration of the environment and planning for rearrangement, which is challenging because of long-horizon nature of the problem. To make progress on this problem, we first analyze the effects of various factors such as number of objects and receptacles, agent carrying capacity, environment layouts etc. on exploration and planning for rearrangement using classical methods. We then investigate both monolithic and modular deep reinforcement learning (DRL) methods for planning in our setting. We find that monolithic DRL methods do not succeed at long-horizon planning needed for multi-object rearrangement. Instead, modular greedy approaches surprisingly perform reasonably well and emerge as competitive baselines for planning with partial observability in multi-object rearrangement problems. We also show that our greedy modular agents are empirically optimal when the objects that need to be rearranged are uniformly distributed in the environment -- thereby contributing baselines with strong performance for future work on multi-object rearrangement planning in partially observable settings.
Large-scale object detection and instance segmentation face a severe data imbalance. The finer-grained object classes become, the less frequent they appear in our datasets. However, at test-time, we expect a detector that performs well for all classes and not just the most frequent ones. In this paper, we provide a theoretical understanding of the long-trail detection problem. We show how the commonly used mean average precision evaluation metric on an unknown test set is bound by a margin-based binary classification error on a long-tailed object detection training set. We optimize margin-based binary classification error with a novel surrogate objective called \textbf{Effective Class-Margin Loss} (ECM). The ECM loss is simple, theoretically well-motivated, and outperforms other heuristic counterparts on LVIS v1 benchmark over a wide range of architecture and detectors. Code is available at \url{https://github.com/janghyuncho/ECM-Loss}.
Object recognition systems are usually trained and evaluated on high resolution images. However, in real world applications, it is common that the images have low resolutions or have small sizes. In this study, we first track the performance of the state-of-the-art deep object recognition network, Faster- RCNN, as a function of image resolution. The results reveals negative effects of low resolution images on recognition performance. They also show that different spatial frequencies convey different information about the objects in recognition process. It means multi-resolution recognition system can provides better insight into optimal selection of features that results in better recognition of objects. This is similar to the mechanisms of the human visual systems that are able to implement multi-scale representation of a visual scene simultaneously. Then, we propose a multi-resolution object recognition framework rather than a single-resolution network. The proposed framework is evaluated on the PASCAL VOC2007 database. The experimental results show the performance of our adapted multi-resolution Faster-RCNN framework outperforms the single-resolution Faster-RCNN on input images with various resolutions with an increase in the mean Average Precision (mAP) of 9.14% across all resolutions and 1.2% on the full-spectrum images. Furthermore, the proposed model yields robustness of the performance over a wide range of spatial frequencies.
Future collaborative robots must be capable of finding objects. As such a fundamental skill, we expect object search to eventually become an off-the-shelf capability for any robot, similar to e.g., object detection, SLAM, and motion planning. However, existing approaches either make unrealistic compromises (e.g., reduce the problem from 3D to 2D), resort to ad-hoc, greedy search strategies, or attempt to learn end-to-end policies in simulation that are yet to generalize across real robots and environments. This thesis argues that through using Partially Observable Markov Decision Processes (POMDPs) to model object search while exploiting structures in the human world (e.g., octrees, correlations) and in human-robot interaction (e.g., spatial language), a practical and effective system for generalized object search can be achieved. In support of this argument, I develop methods and systems for (multi-)object search in 3D environments under uncertainty due to limited field of view, occlusion, noisy, unreliable detectors, spatial correlations between objects, and possibly ambiguous spatial language (e.g., "The red car is behind Chase Bank"). Besides evaluation in simulators such as PyGame, AirSim, and AI2-THOR, I design and implement a robot-independent, environment-agnostic system for generalized object search in 3D and deploy it on the Boston Dynamics Spot robot, the Kinova MOVO robot, and the Universal Robots UR5e robotic arm, to perform object search in different environments. The system enables, for example, a Spot robot to find a toy cat hidden underneath a couch in a kitchen area in under one minute. This thesis also broadly surveys the object search literature, proposing taxonomies in object search problem settings, methods and systems.
When perceiving the world from multiple viewpoints, humans have the ability to reason about the complete objects in a compositional manner even when an object is completely occluded from certain viewpoints. Meanwhile, humans are able to imagine novel views after observing multiple viewpoints. Recent remarkable advances in multi-view object-centric learning still leaves some unresolved problems: 1) The shapes of partially or completely occluded objects can not be well reconstructed. 2) The novel viewpoint prediction depends on expensive viewpoint annotations rather than implicit rules in view representations. In this paper, we introduce a time-conditioned generative model for videos. To reconstruct the complete shape of an object accurately, we enhance the disentanglement between the latent representations of objects and views, where the latent representations of time-conditioned views are jointly inferred with a Transformer and then are input to a sequential extension of Slot Attention to learn object-centric representations. In addition, Gaussian processes are employed as priors of view latent variables for video generation and novel-view prediction without viewpoint annotations. Experiments on multiple datasets demonstrate that the proposed model can make object-centric video decomposition, reconstruct the complete shapes of occluded objects, and make novel-view predictions.
Contour-based instance segmentation has been actively studied, thanks to its flexibility and elegance in processing visual objects within complex backgrounds. In this work, we propose a novel deep network architecture, i.e., PolySnake, for generic contour-based instance segmentation. Motivated by the classic Snake algorithm, the proposed PolySnake achieves superior and robust segmentation performance with an iterative and progressive contour refinement strategy. Technically, PolySnake introduces a recurrent update operator to estimate the object contour iteratively. It maintains a single estimate of the contour that is progressively deformed toward the object boundary. At each iteration, PolySnake builds a semantic-rich representation for the current contour and feeds it to the recurrent operator for further contour adjustment. Through the iterative refinements, the contour progressively converges to a stable status that tightly encloses the object instance. Beyond the scope of general instance segmentation, extensive experiments are conducted to validate the effectiveness and generalizability of our PolySnake in two additional specific task scenarios, including scene text detection and lane detection. The results demonstrate that the proposed PolySnake outperforms the existing advanced methods on several multiple prevalent benchmarks across the three tasks. The codes and pre-trained models are available at https://github.com/fh2019ustc/PolySnake
Active perception describes a broad class of techniques that couple planning and perception systems to move the robot in a way to give the robot more information about the environment. In most robotic systems, perception is typically independent of motion planning. For example, traditional object detection is passive: it operates only on the images it receives. However, we have a chance to improve the results if we allow planning to consume detection signals and move the robot to collect views that maximize the quality of the results. In this paper, we use reinforcement learning (RL) methods to control the robot in order to obtain images that maximize the detection quality. Specifically, we propose using a Decision Transformer with online fine-tuning, which first optimizes the policy with a pre-collected expert dataset and then improves the learned policy by exploring better solutions in the environment. We evaluate the performance of proposed method on an interactive dataset collected from an indoor scenario simulator. Experimental results demonstrate that our method outperforms all baselines, including expert policy and pure offline RL methods. We also provide exhaustive analyses of the reward distribution and observation space.
In this paper, we consider the problem of simultaneously detecting objects and inferring their visual attributes in an image, even for those with no manual annotations provided at the training stage, resembling an open-vocabulary scenario. To achieve this goal, we make the following contributions: (i) we start with a naive two-stage approach for open-vocabulary object detection and attribute classification, termed CLIP-Attr. The candidate objects are first proposed with an offline RPN and later classified for semantic category and attributes; (ii) we combine all available datasets and train with a federated strategy to finetune the CLIP model, aligning the visual representation with attributes, additionally, we investigate the efficacy of leveraging freely available online image-caption pairs under weakly supervised learning; (iii) in pursuit of efficiency, we train a Faster-RCNN type model end-to-end with knowledge distillation, that performs class-agnostic object proposals and classification on semantic categories and attributes with classifiers generated from a text encoder; Finally, (iv) we conduct extensive experiments on VAW, MS-COCO, LSA, and OVAD datasets, and show that recognition of semantic category and attributes is complementary for visual scene understanding, i.e., jointly training object detection and attributes prediction largely outperform existing approaches that treat the two tasks independently, demonstrating strong generalization ability to novel attributes and categories.
Autonomous navigation and path-planning around non-cooperative space objects is an enabling technology for on-orbit servicing and space debris removal systems. The navigation task includes the determination of target object motion, the identification of target object features suitable for grasping, and the identification of collision hazards and other keep-out zones. Given this knowledge, chaser spacecraft can be guided towards capture locations without damaging the target object or without unduly the operations of a servicing target by covering up solar arrays or communication antennas. One way to autonomously achieve target identification, characterization and feature recognition is by use of artificial intelligence algorithms. This paper discusses how the combination of cameras and machine learning algorithms can achieve the relative navigation task. The performance of two deep learning-based object detection algorithms, Faster Region-based Convolutional Neural Networks (R-CNN) and You Only Look Once (YOLOv5), is tested using experimental data obtained in formation flight simulations in the ORION Lab at Florida Institute of Technology. The simulation scenarios vary the yaw motion of the target object, the chaser approach trajectory, and the lighting conditions in order to test the algorithms in a wide range of realistic and performance limiting situations. The data analyzed include the mean average precision metrics in order to compare the performance of the object detectors. The paper discusses the path to implementing the feature recognition algorithms and towards integrating them into the spacecraft Guidance Navigation and Control system.
Images fed to a deep neural network have in general undergone several handcrafted image signal processing (ISP) operations, all of which have been optimized to produce visually pleasing images. In this work, we investigate the hypothesis that the intermediate representation of visually pleasing images is sub-optimal for downstream computer vision tasks compared to the RAW image representation. We suggest that the operations of the ISP instead should be optimized towards the end task, by learning the parameters of the operations jointly during training. We extend previous works on this topic and propose a new learnable operation that enables an object detector to achieve superior performance when compared to both previous works and traditional RGB images. In experiments on the open PASCALRAW dataset, we empirically confirm our hypothesis.
Image segmentation is the task of associating pixels in an image with their respective object class labels. It has a wide range of applications in many industries including healthcare, transportation, robotics, fashion, home improvement, and tourism. Many deep learning-based approaches have been developed for image-level object recognition and pixel-level scene understanding-with the latter requiring a much denser annotation of scenes with a large set of objects. Extensions of image segmentation tasks include 3D and video segmentation, where units of voxels, point clouds, and video frames are classified into different objects. We use "Object Segmentation" to refer to the union of these segmentation tasks. In this monograph, we investigate both traditional and modern object segmentation approaches, comparing their strengths, weaknesses, and utilities. We examine in detail the wide range of deep learning-based segmentation techniques developed in recent years, provide a review of the widely used datasets and evaluation metrics, and discuss potential future research directions.
We propose a new method for object pose estimation without CAD models. The previous feature-matching-based method OnePose has shown promising results under a one-shot setting which eliminates the need for CAD models or object-specific training. However, OnePose relies on detecting repeatable image keypoints and is thus prone to failure on low-textured objects. We propose a keypoint-free pose estimation pipeline to remove the need for repeatable keypoint detection. Built upon the detector-free feature matching method LoFTR, we devise a new keypoint-free SfM method to reconstruct a semi-dense point-cloud model for the object. Given a query image for object pose estimation, a 2D-3D matching network directly establishes 2D-3D correspondences between the query image and the reconstructed point-cloud model without first detecting keypoints in the image. Experiments show that the proposed pipeline outperforms existing one-shot CAD-model-free methods by a large margin and is comparable to CAD-model-based methods on LINEMOD even for low-textured objects. We also collect a new dataset composed of 80 sequences of 40 low-textured objects to facilitate future research on one-shot object pose estimation. The supplementary material, code and dataset are available on the project page: https://zju3dv.github.io/onepose_plus_plus/.
Scene-graph generation involves creating a structural representation of the relationships between objects in a scene by predicting subject-object-relation triplets from input data. Existing methods show poor performance in detecting triplets outside of a predefined set, primarily due to their reliance on dependent feature learning. To address this issue, we propose DDS -- a decoupled dynamic scene-graph generation network -- that consists of two independent branches that can disentangle extracted features. The key innovation of the current paper is the decoupling of the features representing the relationships from those of the objects, which enables the detection of novel object-relationship combinations. The DDS model is evaluated on three datasets and outperforms previous methods by a significant margin, especially in detecting previously unseen triplets.
We construct the first markerless deformable interaction dataset recording interactive motions of the hands and deformable objects, called HMDO (Hand Manipulation with Deformable Objects). With our built multi-view capture system, it captures the deformable interactions with multiple perspectives, various object shapes, and diverse interactive forms. Our motivation is the current lack of hand and deformable object interaction datasets, as 3D hand and deformable object reconstruction is challenging. Mainly due to mutual occlusion, the interaction area is difficult to observe, the visual features between the hand and the object are entangled, and the reconstruction of the interaction area deformation is difficult. To tackle this challenge, we propose a method to annotate our captured data. Our key idea is to collaborate with estimated hand features to guide the object global pose estimation, and then optimize the deformation process of the object by analyzing the relationship between the hand and the object. Through comprehensive evaluation, the proposed method can reconstruct interactive motions of hands and deformable objects with high quality. HMDO currently consists of 21600 frames over 12 sequences. In the future, this dataset could boost the research of learning-based reconstruction of deformable interaction scenes.
Recent advances in modeling 3D objects mostly rely on synthetic datasets due to the lack of large-scale realscanned 3D databases. To facilitate the development of 3D perception, reconstruction, and generation in the real world, we propose OmniObject3D, a large vocabulary 3D object dataset with massive high-quality real-scanned 3D objects. OmniObject3D has several appealing properties: 1) Large Vocabulary: It comprises 6,000 scanned objects in 190 daily categories, sharing common classes with popular 2D datasets (e.g., ImageNet and LVIS), benefiting the pursuit of generalizable 3D representations. 2) Rich Annotations: Each 3D object is captured with both 2D and 3D sensors, providing textured meshes, point clouds, multiview rendered images, and multiple real-captured videos. 3) Realistic Scans: The professional scanners support highquality object scans with precise shapes and realistic appearances. With the vast exploration space offered by OmniObject3D, we carefully set up four evaluation tracks: a) robust 3D perception, b) novel-view synthesis, c) neural surface reconstruction, and d) 3D object generation. Extensive studies are performed on these four benchmarks, revealing new observations, challenges, and opportunities for future research in realistic 3D vision.
Local-HDP (for Local Hierarchical Dirichlet Process) is a hierarchical Bayesian method that has recently been used for open-ended 3D object category recognition. This method has been proven to be efficient in real-time robotic applications. However, the method is not robust to a high degree of occlusion. We address this limitation in two steps. First, we propose a novel semantic 3D object-parts segmentation method that has the flexibility of Local-HDP. This method is shown to be suitable for open-ended scenarios where the number of 3D objects or object parts is not fixed and can grow over time. We show that the proposed method has a higher percentage of mean intersection over union, using a smaller number of learning instances. Second, we integrate this technique with a recently introduced argumentation-based online incremental learning method, thereby enabling the model to handle a high degree of occlusion. We show that the resulting model produces an explicit set of explanations for the 3D object category recognition task.
Object detection and pose estimation are difficult tasks in robotics and autonomous driving. Existing object detection and pose estimation methods mostly adopt the same-dimensional data for training. For example, 2D object detection usually requires a large amount of 2D annotation data with high cost. Using high-dimensional information to supervise lower-dimensional tasks is a feasible way to reduce datasets size. In this work, the DR-WLC, a dimensionality reduction cognitive model, which can perform both object detection and pose estimation tasks at the same time is proposed. The model only requires 3D model of objects and unlabeled environment images (with or without objects) to finish the training. In addition, a bounding boxes generation strategy is also proposed to build the relationship between 3D model and 2D object detection task. Experiments show that our method can qualify the work without any manual annotations and it is easy to deploy for practical applications. Source code is at https://github.com/IN2-ViAUn/DR-WLC.
Imaging and perception in photon-limited scenarios is necessary for various applications, e.g., night surveillance or photography, high-speed photography, and autonomous driving. In these cases, cameras suffer from low signal-to-noise ratio, which degrades the image quality severely and poses challenges for downstream high-level vision tasks like object detection and recognition. Data-driven methods have achieved enormous success in both image restoration and high-level vision tasks. However, the lack of high-quality benchmark dataset with task-specific accurate annotations for photon-limited images/videos delays the research progress heavily. In this paper, we contribute the first multi-illuminance, multi-camera, and low-light dataset, named DarkVision, serving for both image enhancement and object detection. We provide bright and dark pairs with pixel-wise registration, in which the bright counterpart provides reliable reference for restoration and annotation. The dataset consists of bright-dark pairs of 900 static scenes with objects from 15 categories, and 32 dynamic scenes with 4-category objects. For each scene, images/videos were captured at 5 illuminance levels using three cameras of different grades, and average photons can be reliably estimated from the calibration data for quantitative studies. The static-scene images and dynamic videos respectively contain around 7,344 and 320,667 instances in total. With DarkVision, we established baselines for image/video enhancement and object detection by representative algorithms. To demonstrate an exemplary application of DarkVision, we propose two simple yet effective approaches for improving performance in video enhancement and object detection respectively. We believe DarkVision would advance the state-of-the-arts in both imaging and related computer vision tasks in low-light environment.