Pre-trained vision-language models (VLMs) learn to align vision and language representations on large-scale datasets, where each image-text pair usually contains a bag of semantic concepts. However, existing open-vocabulary object detectors only align region embeddings individually with the corresponding features extracted from the VLMs. Such a design leaves the compositional structure of semantic concepts in a scene under-exploited, although the structure may be implicitly learned by the VLMs. In this work, we propose to align the embedding of bag of regions beyond individual regions. The proposed method groups contextually interrelated regions as a bag. The embeddings of regions in a bag are treated as embeddings of words in a sentence, and they are sent to the text encoder of a VLM to obtain the bag-of-regions embedding, which is learned to be aligned to the corresponding features extracted by a frozen VLM. Applied to the commonly used Faster R-CNN, our approach surpasses the previous best results by 4.6 box AP50 and 2.8 mask AP on novel categories of open-vocabulary COCO and LVIS benchmarks, respectively. Code and models are available at https://github.com/wusize/ovdet.
In this paper, we introduce the spatial bias to learn global knowledge without self-attention in convolutional neural networks. Owing to the limited receptive field, conventional convolutional neural networks suffer from learning long-range dependencies. Non-local neural networks have struggled to learn global knowledge, but unavoidably have too heavy a network design due to the self-attention operation. Therefore, we propose a fast and lightweight spatial bias that efficiently encodes global knowledge without self-attention on convolutional neural networks. Spatial bias is stacked on the feature map and convolved together to adjust the spatial structure of the convolutional features. Therefore, we learn the global knowledge on the convolution layer directly with very few additional resources. Our method is very fast and lightweight due to the attention-free non-local method while improving the performance of neural networks considerably. Compared to non-local neural networks, the spatial bias use about 10 times fewer parameters while achieving comparable performance with 1.6 ~ 3.3 times more throughput on a very little budget. Furthermore, the spatial bias can be used with conventional non-local neural networks to further improve the performance of the backbone model. We show that the spatial bias achieves competitive performance that improves the classification accuracy by +0.79% and +1.5% on ImageNet-1K and cifar100 datasets. Additionally, we validate our method on the MS-COCO and ADE20K datasets for downstream tasks involving object detection and semantic segmentation.
With the rapid development of intelligent transportation system applications, a tremendous amount of multi-view video data has emerged to enhance vehicle perception. However, performing video analytics efficiently by exploiting the spatial-temporal redundancy from video data remains challenging. Accordingly, we propose a novel traffic-related framework named CEVAS to achieve efficient object detection using multi-view video data. Briefly, a fine-grained input filtering policy is introduced to produce a reasonable region of interest from the captured images. Also, we design a sharing object manager to manage the information of objects with spatial redundancy and share their results with other vehicles. We further derive a content-aware model selection policy to select detection methods adaptively. Experimental results show that our framework significantly reduces response latency while achieving the same detection accuracy as the state-of-the-art methods.
Open-world object detection (OWOD) is a challenging problem that combines object detection with incremental learning and open-set learning. Compared to standard object detection, the OWOD setting is task to: 1) detect objects seen during training while identifying unseen classes, and 2) incrementally learn the knowledge of the identified unknown objects when the corresponding annotations is available. We propose a novel and efficient OWOD solution from a prototype perspective, which we call OCPL: Open-world object detection via discriminative Class Prototype Learning, which consists of a Proposal Embedding Aggregator (PEA), an Embedding Space Compressor (ESC) and a Cosine Similarity-based Classifier (CSC). All our proposed modules aim to learn the discriminative embeddings of known classes in the feature space to minimize the overlapping distributions of known and unknown classes, which is beneficial to differentiate known and unknown classes. Extensive experiments performed on PASCAL VOC and MS-COCO benchmark demonstrate the effectiveness of our proposed method.
Transparent object perception is a crucial skill for applications such as robot manipulation in household and laboratory settings. Existing methods utilize RGB-D or stereo inputs to handle a subset of perception tasks including depth and pose estimation. However, transparent object perception remains to be an open problem. In this paper, we forgo the unreliable depth map from RGB-D sensors and extend the stereo based method. Our proposed method, MVTrans, is an end-to-end multi-view architecture with multiple perception capabilities, including depth estimation, segmentation, and pose estimation. Additionally, we establish a novel procedural photo-realistic dataset generation pipeline and create a large-scale transparent object detection dataset, Syn-TODD, which is suitable for training networks with all three modalities, RGB-D, stereo and multi-view RGB. Project Site: https://ac-rad.github.io/MVTrans/
Deep neural networks (DNNs) have great potential to solve many real-world problems, but they usually require an extensive amount of computation and memory. It is of great difficulty to deploy a large DNN model to a single resource-limited device with small memory capacity. Distributed computing is a common approach to reduce single-node memory consumption and to accelerate the inference of DNN models. In this paper, we explore the "within-layer model parallelism", which distributes the inference of each layer into multiple nodes. In this way, the memory requirement can be distributed to many nodes, making it possible to use several edge devices to infer a large DNN model. Due to the dependency within each layer, data communications between nodes during this parallel inference can be a bottleneck when the communication bandwidth is limited. We propose a framework to train DNN models for Distributed Inference with Sparse Communications (DISCO). We convert the problem of selecting which subset of data to transmit between nodes into a model optimization problem, and derive models with both computation and communication reduction when each layer is inferred on multiple nodes. We show the benefit of the DISCO framework on a variety of CV tasks such as image classification, object detection, semantic segmentation, and image super resolution. The corresponding models include important DNN building blocks such as convolutions and transformers. For example, each layer of a ResNet-50 model can be distributively inferred across two nodes with five times less data communications, almost half overall computations and half memory requirement for a single node, and achieve comparable accuracy to the original ResNet-50 model. This also results in 4.7 times overall inference speedup.
Sensor fusion is an essential topic in many perception systems, such as autonomous driving and robotics. Transformers-based detection head and CNN-based feature encoder to extract features from raw sensor-data has emerged as one of the best performing sensor-fusion 3D-detection-framework, according to the dataset leaderboards. In this work we provide an in-depth literature survey of transformer based 3D-object detection task in the recent past, primarily focusing on the sensor fusion. We also briefly go through the Vision transformers (ViT) basics, so that readers can easily follow through the paper. Moreover, we also briefly go through few of the non-transformer based less-dominant methods for sensor fusion for autonomous driving. In conclusion we summarize with sensor-fusion trends to follow and provoke future research. More updated summary can be found at: https://github.com/ApoorvRoboticist/Transformers-Sensor-Fusion
High-dynamic range (HDR) images are circulated rapidly over the internet with risks of being exploited for unauthorized usage. To protect these images, some HDR image based watermarking (HDR-IW) methods were put forward. However, they inherited the same problem faced by conventional IW methods for standard dynamic range (SDR) images, where only trade-offs among conflicting requirements are managed instead of simultaneous improvement. In this paper, a novel saliency (eye-catching object) detection based trade-off independent HDR-IW is proposed, to simultaneously improve robustness, imperceptibility and payload. First, the host image goes through our proposed salient object detection model to produce a saliency map, which is, in turn, exploited to segment the foreground and background of the host image. Next, the binary watermark is partitioned into the foregrounds and backgrounds using the same mask and scrambled using a random permutation algorithm. Finally, the watermark segments are embedded into selected bit-plane of the corresponding host segments using quantized indexed modulation. Experimental results suggest that the proposed work outperforms state-of-the-art methods in terms of improving the conflicting requirements.
Pre-trained deep image representations are useful for post-training tasks such as classification through transfer learning, image retrieval, and object detection. Data augmentations are a crucial aspect of pre-training robust representations in both supervised and self-supervised settings. Data augmentations explicitly or implicitly promote invariance in the embedding space to the input image transformations. This invariance reduces generalization to those downstream tasks which rely on sensitivity to these particular data augmentations. In this paper, we propose a method of learning representations that are instead equivariant to data augmentations. We achieve this equivariance through the use of steerable representations. Our representations can be manipulated directly in embedding space via learned linear maps. We demonstrate that our resulting steerable and equivariant representations lead to better performance on transfer learning and robustness: e.g. we improve linear probe top-1 accuracy by between 1% to 3% for transfer; and ImageNet-C accuracy by upto 3.4%. We further show that the steerability of our representations provides significant speedup (nearly 50x) for test-time augmentations; by applying a large number of augmentations for out-of-distribution detection, we significantly improve OOD AUC on the ImageNet-C dataset over an invariant representation.
Semi-supervised object detection (SSOD) is a research hot spot in computer vision, which can greatly reduce the requirement for expensive bounding-box annotations. Despite great success, existing progress mainly focuses on two-stage detection networks like FasterRCNN, while the research on one-stage detectors is often ignored. In this paper, we focus on the semi-supervised learning for the advanced and popular one-stage detection network YOLOv5. Compared with Faster-RCNN, the implementation of YOLOv5 is much more complex, and the various training techniques used in YOLOv5 can also reduce the benefit of SSOD. In addition to this challenge, we also reveal two key issues in one-stage SSOD, which are low-quality pseudo-labeling and multi-task optimization conflict, respectively. To address these issues, we propose a novel teacher-student learning recipe called OneTeacher with two innovative designs, namely Multi-view Pseudo-label Refinement (MPR) and Decoupled Semi-supervised Optimization (DSO). In particular, MPR improves the quality of pseudo-labels via augmented-view refinement and global-view filtering, and DSO handles the joint optimization conflicts via structure tweaks and task-specific pseudo-labeling. In addition, we also carefully revise the implementation of YOLOv5 to maximize the benefits of SSOD, which is also shared with the existing SSOD methods for fair comparison. To validate OneTeacher, we conduct extensive experiments on COCO and Pascal VOC. The extensive experiments show that OneTeacher can not only achieve superior performance than the compared methods, e.g., 15.0% relative AP gains over Unbiased Teacher, but also well handle the key issues in one-stage SSOD. Our source code is available at: https://github.com/luogen1996/OneTeacher.
Fully supervised salient object detection (SOD) methods have made considerable progress in performance, yet these models rely heavily on expensive pixel-wise labels. Recently, to achieve a trade-off between labeling burden and performance, scribble-based SOD methods have attracted increasing attention. Previous scribble-based models directly implement the SOD task only based on SOD training data with limited information, it is extremely difficult for them to understand the image and further achieve a superior SOD task. In this paper, we propose a simple yet effective framework guided by general visual representations with rich contextual semantic knowledge for scribble-based SOD. These general visual representations are generated by self-supervised learning based on large-scale unlabeled datasets. Our framework consists of a task-related encoder, a general visual module, and an information integration module to efficiently combine the general visual representations with task-related features to perform the SOD task based on understanding the contextual connections of images. Meanwhile, we propose a novel global semantic affinity loss to guide the model to perceive the global structure of the salient objects. Experimental results on five public benchmark datasets demonstrate that our method, which only utilizes scribble annotations without introducing any extra label, outperforms the state-of-the-art weakly supervised SOD methods. Specifically, it outperforms the previous best scribble-based method on all datasets with an average gain of 5.5% for max f-measure, 5.8% for mean f-measure, 24% for MAE, and 3.1% for E-measure. Moreover, our method achieves comparable or even superior performance to the state-of-the-art fully supervised models.
Object detectors often experience a drop in performance when new environmental conditions are insufficiently represented in the training data. This paper studies how to automatically fine-tune a pre-existing object detector while exploring and acquiring images in a new environment without relying on human intervention, i.e., in a self-supervised fashion. In our setting, an agent initially explores the environment using a pre-trained off-the-shelf detector to locate objects and associate pseudo-labels. By assuming that pseudo-labels for the same object must be consistent across different views, we devise a novel mechanism for producing refined predictions from the consensus among observations. Our approach improves the off-the-shelf object detector by 2.66% in terms of mAP and outperforms the current state of the art without relying on ground-truth annotations.
Monocular 3D object detection reveals an economical but challenging task in autonomous driving. Recently center-based monocular methods have developed rapidly with a great trade-off between speed and accuracy, where they usually depend on the object center's depth estimation via 2D features. However, the visual semantic features without sufficient pixel geometry information, may affect the performance of clues for spatial 3D detection tasks. To alleviate this, we propose MonoPGC, a novel end-to-end Monocular 3D object detection framework with rich Pixel Geometry Contexts. We introduce the pixel depth estimation as our auxiliary task and design depth cross-attention pyramid module (DCPM) to inject local and global depth geometry knowledge into visual features. In addition, we present the depth-space-aware transformer (DSAT) to integrate 3D space position and depth-aware features efficiently. Besides, we design a novel depth-gradient positional encoding (DGPE) to bring more distinct pixel geometry contexts into the transformer for better object detection. Extensive experiments demonstrate that our method achieves the state-of-the-art performance on the KITTI dataset.
Multi-view radar-camera fused 3D object detection provides a farther detection range and more helpful features for autonomous driving, especially under adverse weather. The current radar-camera fusion methods deliver kinds of designs to fuse radar information with camera data. However, these fusion approaches usually adopt the straightforward concatenation operation between multi-modal features, which ignores the semantic alignment with radar features and sufficient correlations across modals. In this paper, we present MVFusion, a novel Multi-View radar-camera Fusion method to achieve semantic-aligned radar features and enhance the cross-modal information interaction. To achieve so, we inject the semantic alignment into the radar features via the semantic-aligned radar encoder (SARE) to produce image-guided radar features. Then, we propose the radar-guided fusion transformer (RGFT) to fuse our radar and image features to strengthen the two modals' correlation from the global scope via the cross-attention mechanism. Extensive experiments show that MVFusion achieves state-of-the-art performance (51.7% NDS and 45.3% mAP) on the nuScenes dataset. We shall release our code and trained networks upon publication.
Oriented object detection is one of the most fundamental and challenging tasks in remote sensing, aiming to locate and classify objects with arbitrary orientations. Recent advancements in deep learning have significantly enhanced the capabilities of oriented object detection. Given the rapid development of this field, this paper presents a comprehensive survey of recent advances in oriented object detection. To be specific, we begin by tracing the technical evolution from horizontal object detection to oriented object detection and highlighting the specific challenges, including feature misalignment, spatial misalignment, and oriented bounding box (OBB) regression problems. Subsequently, we further categorize existing methods into detection framework, OBB regression, and feature representations, and provide an in-depth discussion on how these approaches address the above challenges. In addition, we cover several publicly available datasets and evaluation protocols. Furthermore, we provide a comprehensive comparison and analysis of state-of-the-art methods. Toward the end of this paper, we identify several future directions for oriented object detection.
Multi-view imaging systems enable uniform coverage of 3D space and reduce the impact of occlusion, which is beneficial for 3D object detection and tracking accuracy. However, existing imaging systems built with multi-view cameras or depth sensors are limited by the small applicable scene and complicated composition. In this paper, we propose a wireless multi-view multi-modal 3D imaging system generally applicable to large outdoor scenes, which consists of a master node and several slave nodes. Multiple spatially distributed slave nodes equipped with cameras and LiDARs are connected to form a wireless sensor network. While providing flexibility and scalability, the system applies automatic spatio-temporal calibration techniques to obtain accurate 3D multi-view multi-modal data. This system is the first imaging system that integrates mutli-view RGB cameras and LiDARs in large outdoor scenes among existing 3D imaging systems. We perform point clouds based 3D object detection and long-term tracking using the 3D imaging dataset collected by this system. The experimental results show that multi-view point clouds greatly improve 3D object detection and tracking accuracy regardless of complex and various outdoor environments.
Automotive radar has increasingly attracted attention due to growing interest in autonomous driving technologies. Acquiring situational awareness using multimodal data collected at high sampling rates by various sensing devices including cameras, LiDAR, and radar requires considerable power, memory and compute resources which are often limited at an edge device. In this paper, we present a novel adaptive radar sub-sampling algorithm designed to identify regions that require more detailed/accurate reconstruction based on prior environmental conditions' knowledge, enabling near-optimal performance at considerably lower effective sampling rates. Designed to robustly perform under variable weather conditions, the algorithm was shown on the Oxford raw radar and RADIATE dataset to achieve accurate reconstruction utilizing only 10% of the original samples in good weather and 20% in extreme (snow, fog) weather conditions. A further modification of the algorithm incorporates object motion to enable reliable identification of important regions. This includes monitoring possible future occlusions caused by objects detected in the present frame. Finally, we train a YOLO network on the RADIATE dataset to perform object detection directly on RADAR data and obtain a 6.6% AP50 improvement over the baseline Faster R-CNN network.
To ensure reliable object detection in autonomous systems, the detector must be able to adapt to changes in appearance caused by environmental factors such as time of day, weather, and seasons. Continually adapting the detector to incorporate these changes is a promising solution, but it can be computationally costly. Our proposed approach is to selectively adapt the detector only when necessary, using new data that does not have the same distribution as the current training data. To this end, we investigate three popular metrics for domain gap evaluation and find that there is a correlation between the domain gap and detection accuracy. Therefore, we apply the domain gap as a criterion to decide when to adapt the detector. Our experiments show that our approach has the potential to improve the efficiency of the detector's operation in real-world scenarios, where environmental conditions change in a cyclical manner, without sacrificing the overall performance of the detector. Our code is publicly available at https://github.com/dadung/DGE-CDA.
In this paper, we optimize a faster region-based convolutional neural network (FRCNN) for 1-dimensional (1D) signal processing and electromagnetic spectrum sensing. We target a cluttered radio frequency (RF) environment, where multiple RF transmission can be present at various frequencies with different bandwidths. The challenge is to accurately and quickly detect and localize each signal with minimal prior information of the signal within a band of interest. As the number of wireless devices grow, and devices become more complex from advances such as software defined radio (SDR), this task becomes increasingly difficult. It is important for sensing devices to keep up with this change, to ensure optimal spectrum usage, to monitor traffic over-the-air for security concerns, and for identifying devices in electronic warfare. Machine learning object detection has shown to be effective for spectrum sensing, however current techniques can be slow and use excessive resources. FRCNN has been applied to perform spectrum sensing using 2D spectrograms, however is unable to be applied directly to 1D signals. We optimize FRCNN to handle 1D signals, including fast Fourier transform (FFT) for spectrum sensing. Our results show that our method has better localization performance, and is faster than the 2D equivalent. Additionally, we show a use case where the modulation type of multiple uncooperative transmissions is identified. Finally, we prove our method generalizes to real world scenarios, by testing it over-the-air using SDR.
The field of object detection using Deep Learning (DL) is constantly evolving with many new techniques and models being proposed. YOLOv7 is a state-of-the-art object detector based on the YOLO family of models which have become popular for industrial applications. One such possible application domain can be semiconductor defect inspection. The performance of any machine learning model depends on its hyperparameters. Furthermore, combining predictions of one or more models in different ways can also affect performance. In this research, we experiment with YOLOv7, a recently proposed, state-of-the-art object detector, by training and evaluating models with different hyperparameters to investigate which ones improve performance in terms of detection precision for semiconductor line space pattern defects. The base YOLOv7 model with default hyperparameters and Non Maximum Suppression (NMS) prediction combining outperforms all RetinaNet models from previous work in terms of mean Average Precision (mAP). We find that vertically flipping images randomly during training yields a 3% improvement in the mean AP of all defect classes. Other hyperparameter values improved AP only for certain classes compared to the default model. Combining models that achieve the best AP for different defect classes was found to be an effective ensembling strategy. Combining predictions from ensembles using Weighted Box Fusion (WBF) prediction gave the best performance. The best ensemble with WBF improved on the mAP of the default model by 10%.