When supervising an object detector with weakly labeled data, most existing approaches are prone to trapping in the discriminative object parts, e.g., finding the face of a cat instead of the full body, due to lacking the supervision on the extent of full objects. To address this challenge, we incorporate object segmentation into the detector training, which guides the model to correctly localize the full objects. We propose the multiple instance curriculum learning (MICL) method, which injects curriculum learning (CL) into the multiple instance learning (MIL) framework. The MICL method starts by automatically picking the easy training examples, where the extent of the segmentation masks agree with detection bounding boxes. The training set is gradually expanded to include harder examples to train strong detectors that handle complex images. The proposed MICL method with segmentation in the loop outperforms the state-of-the-art weakly supervised object detectors by a substantial margin on the PASCAL VOC datasets.
3D object detection is an essential task in autonomous driving. Recent techniques excel with highly accurate detection rates, provided the 3D input data is obtained from precise but expensive LiDAR technology. Approaches based on cheaper monocular or stereo imagery data have, until now, resulted in drastically lower accuracies --- a gap that is commonly attributed to poor image-based depth estimation. However, in this paper we argue that it is not the quality of the data but its representation that accounts for the majority of the difference. Taking the inner workings of convolutional neural networks into consideration, we propose to convert image-based depth maps to pseudo-LiDAR representations --- essentially mimicking the LiDAR signal. With this representation we can apply different existing LiDAR-based detection algorithms. On the popular KITTI benchmark, our approach achieves impressive improvements over the existing state-of-the-art in image-based performance --- raising the detection accuracy of objects within the 30m range from the previous state-of-the-art of 22% to an unprecedented 74%. At the time of submission our algorithm holds the highest entry on the KITTI 3D object detection leaderboard for stereo-image-based approaches. Our code is publicly available at https://github.com/mileyan/pseudo_lidar.
The most common paradigm for vision-based multi-object tracking is tracking-by-detection, due to the availability of reliable detectors for several important object categories such as cars and pedestrians. However, future mobile systems will need a capability to cope with rich human-made environments, in which obtaining detectors for every possible object category would be infeasible. In this paper, we propose a model-free multi-object tracking approach that uses a category-agnostic image segmentation method to track objects. We present an efficient segmentation mask-based tracker which associates pixel-precise masks reported by the segmentation. Our approach can utilize semantic information whenever it is available for classifying objects at the track level, while retaining the capability to track generic unknown objects in the absence of such information. We demonstrate experimentally that our approach achieves performance comparable to state-of-the-art tracking-by-detection methods for popular object categories such as cars and pedestrians. Additionally, we show that the proposed method can discover and robustly track a large variety of other objects.
Fully autonomous driving systems require fast detection and recognition of sensitive objects in the environment. In this context, intelligent vehicles should share their sensor data with computing platforms and/or other vehicles, to detect objects beyond their own sensors' fields of view. However, the resulting huge volumes of data to be exchanged can be challenging to handle for standard communication technologies. In this paper, we evaluate how using a combination of different sensors affects the detection of the environment in which the vehicles move and operate. The final objective is to identify the optimal setup that would minimize the amount of data to be distributed over the channel, with negligible degradation in terms of object detection accuracy. To this aim, we extend an already available object detection algorithm so that it can consider, as an input, camera images, LiDAR point clouds, or a combination of the two, and compare the accuracy performance of the different approaches using two realistic datasets. Our results show that, although sensor fusion always achieves more accurate detections, LiDAR only inputs can obtain similar results for large objects while mitigating the burden on the channel.
Multi-object tracking (MOT) is a core task in computer vision that involves detecting objects in video frames and associating them across time. The rise of deep learning has significantly advanced MOT, particularly within the tracking-by-detection paradigm, which remains the dominant approach. Advancements in modern deep learning-based methods accelerated in 2022 with the introduction of ByteTrack for tracking-by-detection and MOTR for end-to-end tracking. Our survey provides an in-depth analysis of deep learning-based MOT methods, systematically categorizing tracking-by-detection approaches into five groups: joint detection and embedding, heuristic-based, motion-based, affinity learning, and offline methods. In addition, we examine end-to-end tracking methods and compare them with existing alternative approaches. We evaluate the performance of recent trackers across multiple benchmarks and specifically assess their generality by comparing results across different domains. Our findings indicate that heuristic-based methods achieve state-of-the-art results on densely populated datasets with linear object motion, while deep learning-based association methods, in both tracking-by-detection and end-to-end approaches, excel in scenarios with complex motion patterns.
Common object detection models consist of classification and regression branches, due to different task drivers, these two branches have different sensibility to the features from the same scale level and the same spatial location. The point-based prediction method, which is based on the assumption that the high classification confidence point has the high regression quality, leads to the misalignment problem. Our analysis shows, the problem is further composed of scale misalignment and spatial misalignment specifically. We aim to resolve the phenomenon at minimal cost: a minor adjustment of the head network and a new label assignment method replacing the rigid one. Our experiments show that, compared to the baseline FCOS, a one-stage and anchor-free object detection model, our model consistently get around 3 AP improvement with different backbones, demonstrating both simplicity and efficiency of our method.
Existing pseudo label generation methods for point weakly supervised object detection are inadequate in low data volume and dense object detection tasks. We consider the generation of weakly supervised pseudo labels as the model's sparse output, and propose Sparse Generation as a solution to make pseudo labels sparse. The method employs three processing stages (Mapping, Mask, Regression), constructs dense tensors through the relationship between data and detector model, optimizes three of its parameters, and obtains a sparse tensor, thereby indirectly obtaining higher quality pseudo labels, and addresses the model's density problem on low data volume. Additionally, we propose perspective-based matching, which provides more rational pseudo boxes for prediction missed on instances. In comparison to the SOTA method, on four datasets (MS COCO-val, RSOD, SIMD, Bullet-Hole), the experimental results demonstrated a significant advantage.
LIDAR is one of the most important sensors for Unmanned Ground Vehicles (UGV). Object detection and classification based on lidar point cloud is a key technology for UGV. In object detection and classification, the mutual occlusion between neighboring objects is an important factor affecting the accuracy. In this paper, we consider occlusion as an intrinsic property of the point cloud data. We propose a novel approach that explicitly model the occlusion. The occlusion property is then taken into account in the subsequent classification step. We perform experiments on the KITTI dataset. Experimental results indicate that by utilizing the occlusion property that we modeled, the classifier obtains much better performance.
Comprehensive visual understanding requires detection frameworks that can effectively learn and utilize object interactions while analyzing objects individually. This is the main objective in Human-Object Interaction (HOI) detection task. In particular, relative spatial reasoning and structural connections between objects are essential cues for analyzing interactions, which is addressed by the proposed Visual-Spatial-Graph Network (VSGNet) architecture. VSGNet extracts visual features from the human-object pairs, refines the features with spatial configurations of the pair, and utilizes the structural connections between the pair via graph convolutions. The performance of VSGNet is thoroughly evaluated using the Verbs in COCO (V-COCO) and HICO-DET datasets. Experimental results indicate that VSGNet outperforms state-of-the-art solutions by 8% or 4 mAP in V-COCO and 16% or 3 mAP in HICO-DET.
One-stage object detection is commonly implemented by optimizing two sub-tasks: object classification and localization, using heads with two parallel branches, which might lead to a certain level of spatial misalignment in predictions between the two tasks. In this work, we propose a Task-aligned One-stage Object Detection (TOOD) that explicitly aligns the two tasks in a learning-based manner. First, we design a novel Task-aligned Head (T-Head) which offers a better balance between learning task-interactive and task-specific features, as well as a greater flexibility to learn the alignment via a task-aligned predictor. Second, we propose Task Alignment Learning (TAL) to explicitly pull closer (or even unify) the optimal anchors for the two tasks during training via a designed sample assignment scheme and a task-aligned loss. Extensive experiments are conducted on MS-COCO, where TOOD achieves a 51.1 AP at single-model single-scale testing. This surpasses the recent one-stage detectors by a large margin, such as ATSS (47.7 AP), GFL (48.2 AP), and PAA (49.0 AP), with fewer parameters and FLOPs. Qualitative results also demonstrate the effectiveness of TOOD for better aligning the tasks of object classification and localization. Code is available at https://github.com/fcjian/TOOD.
Most models tasked to ground referential utterances in 2D and 3D scenes learn to select the referred object from a pool of object proposals provided by a pre-trained detector. This is limiting because an utterance may refer to visual entities at various levels of granularity, such as the chair, the leg of the chair, or the tip of the front leg of the chair, which may be missed by the detector. We propose a language grounding model that attends on the referential utterance and on the object proposal pool computed from a pre-trained detector to decode referenced objects with a detection head, without selecting them from the pool. In this way, it is helped by powerful pre-trained object detectors without being restricted by their misses. We call our model Bottom Up Top Down DEtection TRansformers (BUTD-DETR) because it uses both language guidance (top down) and objectness guidance (bottom-up) to ground referential utterances in images and point clouds. Moreover, BUTD-DETR casts object detection as referential grounding and uses object labels as language prompts to be grounded in the visual scene, augmenting supervision for the referential grounding task in this way. The proposed model sets a new state-of-the-art across popular 3D language grounding benchmarks with significant performance gains over previous 3D approaches (12.6% on SR3D, 11.6% on NR3D and 6.3% on ScanRefer). When applied in 2D images, it performs on par with the previous state of the art. We ablate the design choices of our model and quantify their contribution to performance. Our code and checkpoints can be found at the project website https://butd-detr.github.io.
We propose a method which can detect events in videos by modeling the change in appearance of the event participants over time. This method makes it possible to detect events which are characterized not by motion, but by the changing state of the people or objects involved. This is accomplished by using object detectors as output models for the states of a hidden Markov model (HMM). The method allows an HMM to model the sequence of poses of the event participants over time, and is effective for poses of humans and inanimate objects. The ability to use existing object-detection methods as part of an event model makes it possible to leverage ongoing work in the object-detection community. A novel training method uses an EM loop to simultaneously learn the temporal structure and object models automatically, without the need to specify either the individual poses to be modeled or the frames in which they occur. The E-step estimates the latent assignment of video frames to HMM states, while the M-step estimates both the HMM transition probabilities and state output models, including the object detectors, which are trained on the weighted subset of frames assigned to their state. A new dataset was gathered because little work has been done on events characterized by changing object pose, and suitable datasets are not available. Our method produced results superior to that of comparison systems on this dataset.
We study the problem of object detection from a novel perspective in which annotation budget constraints are taken into consideration, appropriately coined Budget Aware Object Detection (BAOD). When provided with a fixed budget, we propose a strategy for building a diverse and informative dataset that can be used to optimally train a robust detector. We investigate both optimization and learning-based methods to sample which images to annotate and what type of annotation (strongly or weakly supervised) to annotate them with. We adopt a hybrid supervised learning framework to train the object detector from both these types of annotation. We conduct a comprehensive empirical study showing that a handcrafted optimization method outperforms other selection techniques including random sampling, uncertainty sampling and active learning. By combining an optimal image/annotation selection scheme with hybrid supervised learning to solve the BAOD problem, we show that one can achieve the performance of a strongly supervised detector on PASCAL-VOC 2007 while saving 12.8% of its original annotation budget. Furthermore, when $100\%$ of the budget is used, it surpasses this performance by 2.0 mAP percentage points.
This paper proposes an efficient and effective scheme to applying the sliding window approach popular in computer vision to 3D data. Specifically, the sparse nature of the problem is exploited via a voting scheme to enable a search through all putative object locations at any orientation. We prove that this voting scheme is mathematically equivalent to a convolution on a sparse feature grid and thus enables the processing, in full 3D, of any point cloud irrespective of the number of vantage points required to construct it. As such it is versatile enough to operate on data from popular 3D laser scanners such as a Velodyne as well as on 3D data obtained from increasingly popular push-broom configurations. Our approach is “embarrassingly parallelisable” and capable of processing a point cloud containing over 100K points at eight orientations in less than 0.5s. For the object classes car, pedestrian and bicyclist the resulting detector achieves best-in-class detection and timing performance relative to prior art on the KITTI dataset as well as compared to another existing 3D object detection approach.
Aerators are essential and crucial auxiliary devices in intensive culture, especially in industrial culture in China. The traditional methods cannot accurately detect abnormal condition of aerators in time. Surveillance cameras are widely used as visual perception modules of the Internet of Things, and then using these widely existing surveillance cameras to realize real-time anomaly detection of aerators is a cost-free and easy-to-promote method. However, it is difficult to develop such an expert system due to some technical and applied challenges, e.g., illumination, occlusion, complex background, etc. To tackle these aforementioned challenges, we propose a real-time expert system based on computer vision technology and existing surveillance cameras for anomaly detection of aerators, which consists of two modules, i.e., object region detection and working state detection. First, it is difficult to detect the working state for some small object regions in whole images, and the time complexity of global feature comparison is also high, so we present an object region detection method based on the region proposal idea. Moreover, we propose a novel algorithm called reference frame Kanade-Lucas-Tomasi (RF-KLT) algorithm for motion feature extraction in fixed regions. Then, we present a dimension reduction method of time series for establishing a feature dataset with obvious boundaries between classes. Finally, we use machine learning algorithms to build the feature classifier. The experimental results in both the actual video dataset and the augmented video dataset show that the accuracy for detecting object region and working state of aerators is 100% and 99.9% respectively, and the detection speed is 77-333 frames per second (FPS) according to the different types of surveillance cameras.
In this paper, we present an implicit feature pyramid network (i-FPN) for object detection. Existing FPNs stack several cross-scale blocks to obtain large receptive field. We propose to use an implicit function, recently introduced in deep equilibrium model (DEQ), to model the transformation of FPN. We develop a residual-like iteration to updates the hidden states efficiently. Experimental results on MS COCO dataset show that i-FPN can significantly boost detection performance compared to baseline detectors with ResNet-50-FPN: +3.4, +3.2, +3.5, +4.2, +3.2 mAP on RetinaNet, Faster-RCNN, FCOS, ATSS and AutoAssign, respectively.
This paper presents a detection-aware pre-training (DAP) approach, which leverages only weakly-labeled classification-style datasets (e.g., ImageNet) for pre-training, but is specifically tailored to benefit object detection tasks. In contrast to the widely used image classification-based pre-training (e.g., on ImageNet), which does not include any location-related training tasks, we transform a classification dataset into a detection dataset through a weakly supervised object localization method based on Class Activation Maps to directly pre-train a detector, making the pre-trained model location-aware and capable of predicting bounding boxes. We show that DAP can outperform the traditional classification pre-training in terms of both sample efficiency and convergence speed in downstream detection tasks including VOC and COCO. In particular, DAP boosts the detection accuracy by a large margin when the number of examples in the downstream task is small.
In this paper, we focus on exploring the robustness of the 3D object detection in point clouds, which has been rarely discussed in existing approaches. We observe two crucial phenomena: 1) the detection accuracy of the hard objects, e.g., Pedestrians, is unsatisfactory, 2) when adding additional noise points, the performance of existing approaches decreases rapidly. To alleviate these problems, a novel TANet is introduced in this paper, which mainly contains a Triple Attention (TA) module, and a Coarse-to-Fine Regression (CFR) module. By considering the channel-wise, point-wise and voxel-wise attention jointly, the TA module enhances the crucial information of the target while suppresses the unstable cloud points. Besides, the novel stacked TA further exploits the multi-level feature attention. In addition, the CFR module boosts the accuracy of localization without excessive computation cost. Experimental results on the validation set of KITTI dataset demonstrate that, in the challenging noisy cases, i.e., adding additional random noisy points around each object,the presented approach goes far beyond state-of-the-art approaches. Furthermore, for the 3D object detection task of the KITTI benchmark, our approach ranks the first place on Pedestrian class, by using the point clouds as the only input. The running speed is around 29 frames per second.
Pretraining on large-scale datasets can boost the performance of object detectors while the annotated datasets for object detection are hard to scale up due to the high labor cost. What we possess are numerous isolated filed-specific datasets, thus, it is appealing to jointly pretrain models across aggregation of datasets to enhance data volume and diversity. In this paper, we propose a strong framework for utilizing Multiple datasets to pretrain DETR-like detectors, termed METR, without the need for manual label spaces integration. It converts the typical multi-classification in object detection into binary classification by introducing a pre-trained language model. Specifically, we design a category extraction module for extracting potential categories involved in an image and assign these categories into different queries by language embeddings. Each query is only responsible for predicting a class-specific object. Besides, to adapt our novel detection paradigm, we propose a group bipartite matching strategy that limits the ground truths to match queries assigned to the same category. Extensive experiments demonstrate that METR achieves extraordinary results on either multi-task joint training or the pretrain & finetune paradigm. Notably, our pre-trained models have high flexible transferability and increase the performance upon various DETR-like detectors on COCO val2017 benchmark. Codes will be available after this paper is published.
In Few-Shot Object Detection (FSOD), detecting small objects is extremely difficult. The limited supervision cripples the localization capabilities of the models and a few pixels shift can dramatically reduce the Intersection over Union (IoU) between the ground truth and predicted boxes for small objects. To this end, we propose Scale-adaptive Intersection over Union (SIoU), a novel box similarity measure. SIoU changes with the objects' size, it is more lenient with small object shifts. We conducted a user study and SIoU better aligns than IoU with human judgment. Employing SIoU as an evaluation criterion helps to build more user-oriented models. SIoU can also be used as a loss function to prioritize small objects during training, outperforming existing loss functions. SIoU improves small object detection in the non-few-shot regime, but this setting is unrealistic in the industry as annotated detection datasets are often too expensive to acquire. Hence, our experiments mainly focus on the few-shot regime to demonstrate the superiority and versatility of SIoU loss. SIoU improves significantly FSOD performance on small objects in both natural (Pascal VOC and COCO datasets) and aerial images (DOTA and DIOR). In aerial imagery, small objects are critical and SIoU loss achieves new state-of-the-art FSOD on DOTA and DIOR.