Collecting and annotating real-world data for the development of object detection models is a time-consuming and expensive process. In the military domain in particular, data collection can also be dangerous or infeasible. Training models on synthetic data may provide a solution for cases where access to real-world training data is restricted. However, bridging the reality gap between synthetic and real data remains a challenge. Existing methods usually build on top of baseline Convolutional Neural Network (CNN) models that have been shown to perform well when trained on real data, but have limited ability to perform well when trained on synthetic data. For example, some architectures allow for fine-tuning with the expectation of large quantities of training data and are prone to overfitting on synthetic data. Related work usually ignores various best practices from object detection on real data, e.g. by training on synthetic data from a single environment with relatively little variation. In this paper we propose a methodology for improving the performance of a pre-trained object detector when training on synthetic data. Our approach focuses on extracting the salient information from synthetic data without forgetting useful features learned from pre-training on real images. Based on the state of the art, we incorporate data augmentation methods and a Transformer backbone. Besides reaching relatively strong performance without any specialized synthetic data transfer methods, we show that our methods improve the state of the art on synthetic data trained object detection for the RarePlanes and DGTA-VisDrone datasets, and reach near-perfect performance on an in-house vehicle detection dataset.
Popular Hough Transform-based object detection approaches usually construct an appearance codebook by clustering local image features. However, how to choose appropriate values for the parameters used in the clustering step remains an open problem. Moreover, some popular histogram features extracted from overlapping image blocks may cause a high degree of redundancy and multicollinearity. In this paper, we propose a novel Hough Transform-based object detection approach. First, to address the above issues, we exploit a Bridge Partial Least Squares (BPLS) technique to establish context-encoded Hough Regression Models (HRMs), which are linear regression models that cast probabilistic Hough votes to predict object locations. BPLS is an efficient variant of Partial Least Squares (PLS). PLS-based regression techniques (including BPLS) can reduce the redundancy and eliminate the multicollinearity of a feature set. And the appropriate value of the only parameter used in PLS (i.e., the number of latent components) can be determined by using a cross-validation procedure. Second, to efficiently handle object scale changes, we propose a novel multi-scale voting scheme. In this scheme, multiple Hough images corresponding to multiple object scales can be obtained simultaneously. Third, an object in a test image may correspond to multiple true and false positive hypotheses at different scales. Based on the proposed multi-scale voting scheme, a principled strategy is proposed to fuse hypotheses to reduce false positives by evaluating normalized pointwise mutual information between hypotheses. In the experiments, we also compare the proposed HRM approach with its several variants to evaluate the influences of its components on its performance. Experimental results show that the proposed HRM approach has achieved desirable performances on popular benchmark datasets.
This paper presents a novel 3D object detection framework that processes LiDAR data directly on its native representation: range images. Benefiting from the compactness of range images, 2D convolutions can efficiently process dense LiDAR data of a scene. To overcome scale sensitivity in this perspective view, a novel range-conditioned dilation (RCD) layer is proposed to dynamically adjust a continuous dilation rate as a function of the measured range. Furthermore, localized soft range gating combined with a 3D box-refinement stage improves robustness in occluded areas, and produces overall more accurate bounding box predictions. On the public large-scale Waymo Open Dataset, our method sets a new baseline for range-based 3D detection, outperforming multiview and voxel-based methods over all ranges with unparalleled performance at long range detection.
In this paper, we study Multiscale Vision Transformers (MViTv2) as a unified architecture for image and video classification, as well as object detection. We present an improved version of MViT that incorporates decomposed relative positional embeddings and residual pooling connections. We instantiate this architecture in five sizes and evaluate it for ImageNet classification, COCO detection and Kinetics video recognition where it outperforms prior work. We further compare MViTv2s' pooling attention to window attention mechanisms where it outperforms the latter in accuracy/compute. Without bells-and-whistles, MViTv2 has state-of-the-art performance in 3 domains: 88.8% accuracy on ImageNet classification, 58.7 boxAP on COCO object detection as well as 86.1% on Kinetics-400 video classification. Code and models are available at https://github.com/facebookresearch/mvit.
Arbitrary-oriented object detection has been a building block for rotation sensitive tasks. We first show that the boundary problem suffered in existing dominant regression-based rotation detectors, is caused by angular periodicity or corner ordering, according to the parameterization protocol. We also show that the root cause is that the ideal predictions can be out of the defined range. Accordingly, we transform the angular prediction task from a regression problem to a classification one. For the resulting circularly distributed angle classification problem, we first devise a Circular Smooth Label technique to handle the periodicity of angle and increase the error tolerance to adjacent angles. To reduce the excessive model parameters by Circular Smooth Label, we further design a Densely Coded Labels, which greatly reduces the length of the encoding. Finally, we further develop an object heading detection module, which can be useful when the exact heading orientation information is needed e.g. for ship and plane heading detection. We release our OHD-SJTU dataset and OHDet detector for heading detection. Extensive experimental results on three large-scale public datasets for aerial images i.e. DOTA, HRSC2016, OHD-SJTU, and face dataset FDDB, as well as scene text dataset ICDAR2015 and MLT, show the effectiveness of our approach.
The performance of object detection, to a great extent, depends on the availability of large annotated datasets. To alleviate the annotation cost, the research community has explored a number of ways to exploit unlabeled or weakly labeled data. However, such efforts have met with limited success so far. In this work, we revisit the problem with a pragmatic standpoint, trying to explore a new balance between detection performance and annotation cost by jointly exploiting fully and weakly annotated data. Specifically, we propose a weakly- and semi-supervised object detection framework (WSSOD), which involves a two-stage learning procedure. An agent detector is first trained on a joint dataset and then used to predict pseudo bounding boxes on weakly-annotated images. The underlying assumptions in the current as well as common semi-supervised pipelines are also carefully examined under a unified EM formulation. On top of this framework, weakly-supervised loss (WSL), label attention and random pseudo-label sampling (RPS) strategies are introduced to relax these assumptions, bringing additional improvement on the efficacy of the detection pipeline. The proposed framework demonstrates remarkable performance on PASCAL-VOC and MSCOCO benchmark, achieving a high performance comparable to those obtained in fully-supervised settings, with only one third of the annotations.
Few-Shot Object Detection (FSOD) methods are mainly designed and evaluated on natural image datasets such as Pascal VOC and MS COCO. However, it is not clear whether the best methods for natural images are also the best for aerial images. Furthermore, direct comparison of performance between FSOD methods is difficult due to the wide variety of detection frameworks and training strategies. Therefore, we propose a benchmarking framework that provides a flexible environment to implement and compare attention-based FSOD methods. The proposed framework focuses on attention mechanisms and is divided into three modules: spatial alignment, global attention, and fusion layer. To remain competitive with existing methods, which often leverage complex training, we propose new augmentation techniques designed for object detection. Using this framework, several FSOD methods are reimplemented and compared. This comparison highlights two distinct performance regimes on aerial and natural images: FSOD performs worse on aerial images. Our experiments suggest that small objects, which are harder to detect in the few-shot setting, account for the poor performance. Finally, we develop a novel multiscale alignment method, Cross-Scales Query-Support Alignment (XQSA) for FSOD, to improve the detection of small objects. XQSA outperforms the state-of-the-art significantly on DOTA and DIOR.
LiDAR-based sensors employing optical spectrum signals play a vital role in providing significant information about the target objects in autonomous driving vehicle systems. However, the presence of fog in the atmosphere severely degrades the overall system's performance. This manuscript analyzes the role of fog particle size distributions in 3D object detection under adverse weather conditions. We utilise Mie theory and meteorological optical range (MOR) to calculate the attenuation and backscattering coefficient values for point cloud generation and analyze the overall system's accuracy in Car, Cyclist, and Pedestrian case scenarios under easy, medium and hard detection difficulties. Gamma and Junge (Power-Law) distributions are employed to mathematically model the fog particle size distribution under strong and moderate advection fog environments. Subsequently, we modified the KITTI dataset based on the backscattering coefficient values and trained it on the PV-RCNN++ deep neural network model for Car, Cyclist, and Pedestrian cases under different detection difficulties. The result analysis shows a significant variation in the system's accuracy concerning the changes in target object dimensionality, the nature of the fog environment and increasing detection difficulties, with the Car exhibiting the highest accuracy of around 99% and the Pedestrian showing the lowest accuracy of around 73%.
The employment of convolutional neural networks has led to significant performance improvement on the task of object detection. However, when applying existing detectors to continuous frames in a video, we often encounter momentary miss-detection of objects, that is, objects are undetected exceptionally at a few frames, although they are correctly detected at all other frames. In this paper, we analyze the mechanism of how such miss-detection occurs. For the most popular class of detectors that are based on anchor boxes, we show the followings: i) besides apparent causes such as motion blur, occlusions, background clutters, etc., the majority of remaining miss-detection can be explained by an improper behavior of the detectors at boundaries of the anchor boxes; and ii) this can be rectified by improving the way of choosing positive samples from candidate anchor boxes when training the detectors.
Searching for objects in unfamiliar scenarios is a challenging task for blind people. It involves specifying the target object, detecting it, and then gathering detailed information according to the user's intent. However, existing description- and detection-based assistive technologies do not sufficiently support the multifaceted nature of interactive object search tasks. We present ObjectFinder, an open-vocabulary wearable assistive system for interactive object search by blind people. ObjectFinder allows users to query target objects using flexible wording. Once the target object is detected, it provides egocentric localization information in real-time, including distance and direction. Users can then initiate different branches to gather detailed information based on their intent towards the target object, such as navigating to it or perceiving its surroundings. ObjectFinder is powered by a seamless combination of open-vocabulary models, namely an open-vocabulary object detector and a multimodal large language model. The ObjectFinder design concept and its development were carried out in collaboration with a blind co-designer. To evaluate ObjectFinder, we conducted an exploratory user study with eight blind participants. We compared ObjectFinder to BeMyAI and Google Lookout, popular description- and detection-based assistive applications. Our findings indicate that most participants felt more independent with ObjectFinder and preferred it for object search, as it enhanced scene context gathering and navigation, and allowed for active target identification. Finally, we discuss the implications for future assistive systems to support interactive object search.
Multi-label image classification is a fundamental but challenging task towards general visual understanding. Existing methods found the region-level cues (e.g., features from RoIs) can facilitate multi-label classification. Nevertheless, such methods usually require laborious object-level annotations (i.e., object labels and bounding boxes) for effective learning of the object-level visual features. In this paper, we propose a novel and efficient deep framework to boost multi-label classification by distilling knowledge from weakly-supervised detection task without bounding box annotations. Specifically, given the image-level annotations, (1) we first develop a weakly-supervised detection (WSD) model, and then (2) construct an end-to-end multi-label image classification framework augmented by a knowledge distillation module that guides the classification model by the WSD model according to the class-level predictions for the whole image and the object-level visual features for object RoIs. The WSD model is the teacher model and the classification model is the student model. After this cross-task knowledge distillation, the performance of the classification model is significantly improved and the efficiency is maintained since the WSD model can be safely discarded in the test phase. Extensive experiments on two large-scale datasets (MS-COCO and NUS-WIDE) show that our framework achieves superior performances over the state-of-the-art methods on both performance and efficiency.
Monocular 3D object detection is a fundamental but very important task to many applications including autonomous driving, robotic grasping and augmented reality. Existing leading methods tend to estimate the depth of the input image first, and detect the 3D object based on point cloud. This routine suffers from the inherent gap between depth estimation and object detection. Besides, the prediction error accumulation would also affect the performance. In this paper, a novel method named MonoSIM is proposed. The insight behind introducing MonoSIM is that we propose to simulate the feature learning behaviors of a point cloud based detector for monocular detector during the training period. Hence, during inference period, the learned features and prediction would be similar to the point cloud based detector as possible. To achieve it, we propose one scene-level simulation module, one RoI-level simulation module and one response-level simulation module, which are progressively used for the detector's full feature learning and prediction pipeline. We apply our method to the famous M3D-RPN detector and CaDDN detector, conducting extensive experiments on KITTI and Waymo Open datasets. Results show that our method consistently improves the performance of different monocular detectors for a large margin without changing their network architectures. Our codes will be publicly available at https://github.com/sunh18/MonoSIM}{https://github.com/sunh18/MonoSIM.
In vision-enabled autonomous systems such as robots and autonomous cars, video object detection plays a crucial role, and both its speed and accuracy are important factors to provide reliable operation. The key insight we show in this paper is that speed and accuracy are not necessarily a trade-off when it comes to image scaling. Our results show that re-scaling the image to a lower resolution will sometimes produce better accuracy. Based on this observation, we propose a novel approach, dubbed AdaScale, which adaptively selects the input image scale that improves both accuracy and speed for video object detection. To this end, our results on ImageNet VID and mini YouTube-BoundingBoxes datasets demonstrate 1.3 points and 2.7 points mAP improvement with 1.6x and 1.8x speedup, respectively. Additionally, we improve state-of-the-art video acceleration work by an extra 1.25x speedup with slightly better mAP on ImageNet VID dataset.
Recent deep learning based salient object detection methods which utilize both saliency and boundary features have achieved remarkable performance. However, most of them ignore the complementarity between saliency features and boundary features, thus get worse predictions in scenes with low contrast between foreground and background. To address this issue, we propose a novel Recurrent Two-Stream Guided Refinement Network (RTGRNet) that consists of iterating Two-Stream Guided Refinement Modules (TGRMs). TGRM consists of a Guide Block and two feature streams: saliency and boundary, the Guide Block utilizes the refined features after previous TGRM to further improve the performance of two feature streams in current TGRM. Meanwhile, the low-level integrated features are also utilized as a reference to get better details. Finally, we progressively refine these features by recurrently stacking more TGRMs. Extensive experiments on six public datasets show that our proposed RTGRNet achieves the state-of-the-art performance in salient object detection.
Confidence calibration is a major concern when applying artificial neural networks in safety-critical applications. Since most research in this area has focused on classification in the past, confidence calibration in the scope of object detection has gained more attention only recently. Based on previous work, we study the miscalibration of object detection models with respect to image location and box scale. Our main contribution is to additionally consider the impact of box selection methods like non-maximum suppression to calibration. We investigate the default intrinsic calibration of object detection models and how it is affected by these post-processing techniques. For this purpose, we distinguish between black-box calibration with non-maximum suppression and white-box calibration with raw network outputs. Our experiments reveal that post-processing highly affects confidence calibration. We show that non-maximum suppression has the potential to degrade initially well-calibrated predictions, leading to overconfident and thus miscalibrated models.
Video salient object detection (VSOD), as a fundamental computer vision problem, has been extensively discussed in the last decade. However, all existing works focus on addressing the VSOD problem in 2D scenarios. With the rapid development of VR devices, panoramic videos have been a promising alternative to 2D videos to provide immersive feelings of the real world. In this paper, we aim to tackle the video salient object detection problem for panoramic videos, with their corresponding ambisonic audios. A multimodal fusion module equipped with two pseudo-siamese audio-visual context fusion (ACF) blocks is proposed to effectively conduct audio-visual interaction. The ACF block equipped with spherical positional encoding enables the fusion in the 3D context to capture the spatial correspondence between pixels and sound sources from the equirectangular frames and ambisonic audios. Experimental results verify the effectiveness of our proposed components and demonstrate that our method achieves state-of-the-art performance on the ASOD60K dataset.
This paper presents how we can achieve the state-of-the-art accuracy in multi-category object detection task while minimizing the computational cost by adapting and combining recent technical innovations. Following the common pipeline of "CNN feature extraction + region proposal + RoI classification", we mainly redesign the feature extraction part, since region proposal part is not computationally expensive and classification part can be efficiently compressed with common techniques like truncated SVD. Our design principle is "less channels with more layers" and adoption of some building blocks including concatenated ReLU, Inception, and HyperNet. The designed network is deep and thin and trained with the help of batch normalization, residual connections, and learning rate scheduling based on plateau detection. We obtained solid results on well-known object detection benchmarks: 83.8% mAP (mean average precision) on VOC2007 and 82.5% mAP on VOC2012 (2nd place), while taking only 750ms/image on Intel i7-6700K CPU with a single core and 46ms/image on NVIDIA Titan X GPU. Theoretically, our network requires only 12.3% of the computational cost compared to ResNet-101, the winner on VOC2012.
Open World Object Detection (OWOD) is a novel computer vision task with a considerable challenge, bridging the gap between classic object detection (OD) benchmarks and real-world object detection. In addition to detecting and classifying seen/known objects, OWOD algorithms are expected to detect unseen/unknown objects and incrementally learn them. The natural instinct of humans to identify unknown objects in their environments mainly depends on their brains' knowledge base. It is difficult for a model to do this only by learning from the annotation of several tiny datasets. The large pre-trained grounded language-image models - VL (\ie GLIP) have rich knowledge about the open world but are limited to the text prompt. We propose leveraging the VL as the ``Brain'' of the open-world detector by simply generating unknown labels. Leveraging it is non-trivial because the unknown labels impair the model's learning of known objects. In this paper, we alleviate these problems by proposing the down-weight loss function and decoupled detection structure. Moreover, our detector leverages the ``Brain'' to learn novel objects beyond VL through our pseudo-labeling scheme.
Deep convolutional neural network significantly boosted the capability of salient object detection in handling large variations of scenes and object appearances. However, convolution operations seek to generate strong responses on individual pixels, while lack the ability to maintain the spatial structure of objects. Moreover, the down-sampling operations, such as pooling and striding, lose spatial details of the salient objects. In this paper, we propose a simple yet effective Siamese Edge-Enhancement Network (SE2Net) to preserve the edge structure for salient object detection. Specifically, a novel multi-stage siamese network is built to aggregate the low-level and high-level features, and parallelly estimate the salient maps of edges and regions. As a result, the predicted regions become more accurate by enhancing the responses at edges, and the predicted edges become more semantic by suppressing the false positives in background. After the refined salient maps of edges and regions are produced by the SE2Net, an edge-guided inference algorithm is designed to further improve the resulting salient masks along the predicted edges. Extensive experiments on several benchmark datasets have been conducted, which show that our method is superior than the state-of-the-art approaches.
In this paper, we propose a simple yet effective approach, named Triple Excitation Network, to reinforce the training of video salient object detection (VSOD) from three aspects, spatial, temporal, and online excitations. These excitation mechanisms are designed following the spirit of curriculum learning and aim to reduce learning ambiguities at the beginning of training by selectively exciting feature activations using ground truth. Then we gradually reduce the weight of ground truth excitations by a curriculum rate and replace it by a curriculum complementary map for better and faster convergence. In particular, the spatial excitation strengthens feature activations for clear object boundaries, while the temporal excitation imposes motions to emphasize spatio-temporal salient regions. Spatial and temporal excitations can combat the saliency shifting problem and conflict between spatial and temporal features of VSOD. Furthermore, our semi-curriculum learning design enables the first online refinement strategy for VSOD, which allows exciting and boosting saliency responses during testing without re-training. The proposed triple excitations can easily plug in different VSOD methods. Extensive experiments show the effectiveness of all three excitation methods and the proposed method outperforms state-of-the-art image and video salient object detection methods.