One of the important bottlenecks in training modern object detectors is the need for labeled images where bounding box annotations have to be produced for each object present in the image. This bottleneck is further exacerbated in aerial images where the annotators have to label small objects often distributed in clusters on high-resolution images. In recent days, the mean-teacher approach trained with pseudo-labels and weak-strong augmentation consistency is gaining popularity for semi-supervised object detection. However, a direct adaptation of such semi-supervised detectors for aerial images where small clustered objects are often present, might not lead to optimal results. In this paper, we propose a density crop-guided semi-supervised detector that identifies the cluster of small objects during training and also exploits them to improve performance at inference. During training, image crops of clusters identified from labeled and unlabeled images are used to augment the training set, which in turn increases the chance of detecting small objects and creating good pseudo-labels for small objects on the unlabeled images. During inference, the detector is not only able to detect the objects of interest but also regions with a high density of small objects (density crops) so that detections from the input image and detections from image crops are combined, resulting in an overall more accurate object prediction, especially for small objects. Empirical studies on the popular benchmarks of VisDrone and DOTA datasets show the effectiveness of our density crop-guided semi-supervised detector with an average improvement of more than 2\% over the basic mean-teacher method in COCO style AP. Our code is available at: //github.com/akhilpm/DroneSSOD.
相關內容
Recently, the remarkable advance of the Large Language Model (LLM) has inspired researchers to transfer its extraordinary reasoning capability to both vision and language data. However, the prevailing approaches primarily regard the visual input as a prompt and focus exclusively on optimizing the text generation process conditioned upon vision content by a frozen LLM. Such an inequitable treatment of vision and language heavily constrains the model's potential. In this paper, we break through this limitation by representing both vision and language in a unified form. Specifically, we introduce a well-designed visual tokenizer to translate the non-linguistic image into a sequence of discrete tokens like a foreign language that LLM can read. The resulting visual tokens encompass high-level semantics worthy of a word and also support dynamic sequence length varying from the image. Coped with this tokenizer, the presented foundation model called LaVIT can handle both image and text indiscriminately under the same generative learning paradigm. This unification empowers LaVIT to serve as an impressive generalist interface to understand and generate multi-modal content simultaneously. Extensive experiments further showcase that it outperforms the existing models by a large margin on massive vision-language tasks. Our code and models will be available at //github.com/jy0205/LaVIT.
Currently, truss tomato weighing and packaging require significant manual work. The main obstacle to automation lies in the difficulty of developing a reliable robotic grasping system for already harvested trusses. We propose a method to grasp trusses that are stacked in a crate with considerable clutter, which is how they are commonly stored and transported after harvest. The method consists of a deep learning-based vision system to first identify the individual trusses in the crate and then determine a suitable grasping location on the stem. To this end, we have introduced a grasp pose ranking algorithm with online learning capabilities. After selecting the most promising grasp pose, the robot executes a pinch grasp without needing touch sensors or geometric models. Lab experiments with a robotic manipulator equipped with an eye-in-hand RGB-D camera showed a 100% clearance rate when tasked to pick all trusses from a pile. 93% of the trusses were successfully grasped on the first try, while the remaining 7% required more attempts.
This report examines the effectiveness of Chain-of-Thought (CoT) prompting in improving the multi-step reasoning abilities of large language models (LLMs). Inspired by previous studies \cite{Min2022RethinkingWork}, we analyze the impact of three types of CoT prompt perturbations, namely CoT order, CoT values, and CoT operators on the performance of GPT-3 on various tasks. Our findings show that incorrect CoT prompting leads to poor performance on accuracy metrics. Correct values in the CoT is crucial for predicting correct answers. Moreover, incorrect demonstrations, where the CoT operators or the CoT order are wrong, do not affect the performance as drastically when compared to the value based perturbations. This research deepens our understanding of CoT prompting and opens some new questions regarding the capability of LLMs to learn reasoning in context.
The primary bottleneck towards obtaining good recognition performance in IR images is the lack of sufficient labeled training data, owing to the cost of acquiring such data. Realizing that object detection methods for the RGB modality are quite robust (at least for some commonplace classes, like person, car, etc.), thanks to the giant training sets that exist, in this work we seek to leverage cues from the RGB modality to scale object detectors to the IR modality, while preserving model performance in the RGB modality. At the core of our method, is a novel tensor decomposition method called TensorFact which splits the convolution kernels of a layer of a Convolutional Neural Network (CNN) into low-rank factor matrices, with fewer parameters than the original CNN. We first pretrain these factor matrices on the RGB modality, for which plenty of training data are assumed to exist and then augment only a few trainable parameters for training on the IR modality to avoid over-fitting, while encouraging them to capture complementary cues from those trained only on the RGB modality. We validate our approach empirically by first assessing how well our TensorFact decomposed network performs at the task of detecting objects in RGB images vis-a-vis the original network and then look at how well it adapts to IR images of the FLIR ADAS v1 dataset. For the latter, we train models under scenarios that pose challenges stemming from data paucity. From the experiments, we observe that: (i) TensorFact shows performance gains on RGB images; (ii) further, this pre-trained model, when fine-tuned, outperforms a standard state-of-the-art object detector on the FLIR ADAS v1 dataset by about 4% in terms of mAP 50 score.
Many libraries, such as OpenCV, FFmpeg, XNNPACK, and Eigen, utilize Arm or x86 SIMD Intrinsics to optimize programs for performance. With the emergence of RISC-V Vector Extensions (RVV), there is a need to migrate these performance legacy codes for RVV. Currently, the migration of NEON code to RVV code requires manual rewriting, which is a time-consuming and error-prone process. In this work, we use the open source tool, "SIMD Everywhere" (SIMDe), to automate the migration. Our primary task is to enhance SIMDe to enable the conversion of ARM NEON Intrinsics types and functions to their corresponding RVV Intrinsics types and functions. For type conversion, we devise strategies to convert Neon Intrinsics types to RVV Intrinsics by considering the vector length agnostic (vla) architectures. With function conversions, we analyze commonly used conversion methods in SIMDe and develop customized conversions for each function based on the results of RVV code generations. In our experiments with Google XNNPACK library, our enhanced SIMDe achieves speedup ranging from 1.51x to 5.13x compared to the original SIMDe, which does not utilize customized RVV implementations for the conversions.
Self-driving software pipelines include components that are learned from a significant number of training examples, yet it remains challenging to evaluate the overall system's safety and generalization performance. Together with scaling up the real-world deployment of autonomous vehicles, it is of critical importance to automatically find simulation scenarios where the driving policies will fail. We propose a method that efficiently generates adversarial simulation scenarios for autonomous driving by solving an optimal control problem that aims to maximally perturb the policy from its nominal trajectory. Given an image-based driving policy, we show that we can inject new objects in a neural rendering representation of the deployment scene, and optimize their texture in order to generate adversarial sensor inputs to the policy. We demonstrate that adversarial scenarios discovered purely in the neural renderer (surrogate scene) can often be successfully transferred to the deployment scene, without further optimization. We demonstrate this transfer occurs both in simulated and real environments, provided the learned surrogate scene is sufficiently close to the deployment scene.
Previous approaches to detecting human anomalies in videos have typically relied on implicit modeling by directly applying the model to video or skeleton data, potentially resulting in inaccurate modeling of motion information. In this paper, we conduct an exploratory study and introduce a new idea called HKVAD (Human Kinematic-inspired Video Anomaly Detection) for video anomaly detection, which involves the explicit use of human kinematic features to detect anomalies. To validate the effectiveness and potential of this perspective, we propose a pilot method that leverages the kinematic features of the skeleton pose, with a specific focus on the walking stride, skeleton displacement at feet level, and neck level. Following this, the method employs a normalizing flow model to estimate density and detect anomalies based on the estimated density. Based on the number of kinematic features used, we have devised three straightforward variant methods and conducted experiments on two highly challenging public datasets, ShanghaiTech and UBnormal. Our method achieves good results with minimal computational resources, validating its effectiveness and potential.
The ability to understand the surrounding scene is of paramount importance for Autonomous Vehicles (AVs). This paper presents a system capable to work in an online fashion, giving an immediate response to the arise of anomalies surrounding the AV, exploiting only the videos captured by a dash-mounted camera. Our architecture, called MOVAD, relies on two main modules: a Short-Term Memory Module to extract information related to the ongoing action, implemented by a Video Swin Transformer (VST), and a Long-Term Memory Module injected inside the classifier that considers also remote past information and action context thanks to the use of a Long-Short Term Memory (LSTM) network. The strengths of MOVAD are not only linked to its excellent performance, but also to its straightforward and modular architecture, trained in a end-to-end fashion with only RGB frames with as less assumptions as possible, which makes it easy to implement and play with. We evaluated the performance of our method on Detection of Traffic Anomaly (DoTA) dataset, a challenging collection of dash-mounted camera videos of accidents. After an extensive ablation study, MOVAD is able to reach an AUC score of 82.17\%, surpassing the current state-of-the-art by +2.87 AUC. Our code will be available on //github.com/IMPLabUniPr/movad/tree/movad_vad
Most object recognition approaches predominantly focus on learning discriminative visual patterns while overlooking the holistic object structure. Though important, structure modeling usually requires significant manual annotations and therefore is labor-intensive. In this paper, we propose to "look into object" (explicitly yet intrinsically model the object structure) through incorporating self-supervisions into the traditional framework. We show the recognition backbone can be substantially enhanced for more robust representation learning, without any cost of extra annotation and inference speed. Specifically, we first propose an object-extent learning module for localizing the object according to the visual patterns shared among the instances in the same category. We then design a spatial context learning module for modeling the internal structures of the object, through predicting the relative positions within the extent. These two modules can be easily plugged into any backbone networks during training and detached at inference time. Extensive experiments show that our look-into-object approach (LIO) achieves large performance gain on a number of benchmarks, including generic object recognition (ImageNet) and fine-grained object recognition tasks (CUB, Cars, Aircraft). We also show that this learning paradigm is highly generalizable to other tasks such as object detection and segmentation (MS COCO). Project page: //github.com/JDAI-CV/LIO.
The recent proliferation of knowledge graphs (KGs) coupled with incomplete or partial information, in the form of missing relations (links) between entities, has fueled a lot of research on knowledge base completion (also known as relation prediction). Several recent works suggest that convolutional neural network (CNN) based models generate richer and more expressive feature embeddings and hence also perform well on relation prediction. However, we observe that these KG embeddings treat triples independently and thus fail to cover the complex and hidden information that is inherently implicit in the local neighborhood surrounding a triple. To this effect, our paper proposes a novel attention based feature embedding that captures both entity and relation features in any given entity's neighborhood. Additionally, we also encapsulate relation clusters and multihop relations in our model. Our empirical study offers insights into the efficacy of our attention based model and we show marked performance gains in comparison to state of the art methods on all datasets.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191