Object detection on drone-captured scenarios is a recent popular task. As drones always navigate in different altitudes, the object scale varies violently, which burdens the optimization of networks. Moreover, high-speed and low-altitude flight bring
in the motion blur on the densely packed objects, which leads to great challenge of object distinction. To solve the two issues mentioned above, we propose TPH-YOLOv5. Based on YOLOv5, we add one more prediction head to detect different-scale objects. Then we replace the original prediction heads with Transformer Prediction Heads (TPH) to explore the prediction potential with self-attention mechanism. We also integrate convolutional block attention model (CBAM) to find attention region on scenarios with dense objects. To achieve more improvement of our proposed TPH-YOLOv5, we provide bags of useful strategies such as data augmentation, multiscale testing, multi-model integration and utilizing extra classifier. Extensive experiments on dataset VisDrone2021 show that TPH-YOLOv5 have good performance with impressive interpretability on drone-captured scenarios. On DET-test-challenge dataset, the AP result of TPH-YOLOv5 are 39.18%, which is better than previous SOTA method (DPNetV3) by 1.81%. On VisDrone Challenge 2021, TPHYOLOv5 wins 5th place and achieves well-matched results with 1st place model (AP 39.43%). Compared to baseline model (YOLOv5), TPH-YOLOv5 improves about 7%, which is encouraging and competitive.
We investigate the frame set of regular multivariate Gaussian Gabor frames using methods from Kahler geometry such as Hormanders $dbar$-method, the Ohsawa--Takegoshi extension theorem and a Kahler-variant of the symplectic embedding theorem of McDuff
-Polterovich for ellipsoids. Our approach is based on the well-known link between sets of interpolation for the Bargmann-Fock space and the frame set of multivariate Gaussian Gabor frames. We state sufficient conditions in terms of a certain extremal type Seshadri constant of the complex torus associated to a lattice to be a set of interpolation for the Bargmann-Fock space, and give also a condition in terms of the generalized Buser-Sarnak invariant of the lattice. Our results on Gaussian Gabor frames are in terms of the Sehsadri constant and the generalized Buser-Sarnak invariant of the associated symplectic dual lattice. The theory of Hormander estimates and the Ohsawa--Takegoshi extension theorem allow us to give estimates for the frame bounds in terms of the Buser-Sarnack invariant and in the one-dimensional case these bounds are sharp thanks to Faltings work on Green functions in Arakelov theory.
As distributed ledgers, blockchains run consensus protocols which trade capacity for consistency, especially in non-ideal networks with incomplete connectivity and erroneous links. Existing studies on the tradeoff between capacity and consistency are
only qualitative or rely on specific assumptions. This paper presents discrete-time Markov chain models to quantify the capacity of Proof-of-Work based public blockchains in non-ideal networks. The comprehensive model is collapsed to be ergodic under the eventual consistency of blockchains, achieving tractability and efficient evaluations of blockchain capacity. A closed-form expression for the capacity is derived in the case of two miners. Another important aspect is that we extend the ergodic model to analyze the capacity under strong consistency, evaluating the robustness of blockchains against double-spending attacks. Validated by simulations, the proposed models are accurate and reveal the effect of link quality and the distribution of mining rates on blockchain capacity and the ratio of stale blocks.
This paper is concerned with an inverse source problem for the stochastic biharmonic operator wave equation. The driven source is assumed to be a microlocally isotropic Gaussian random field with its covariance operator being a classical pseudo-diffe
rential operator. The well-posedness of the direct problem is examined in the distribution sense and the regularity of the solution is discussed for the given rough source. For the inverse problem, the strength of the random source, involved in the principal symbol of its covariance operator, is shown to be uniquely determined by a single realization of the magnitude of the wave field averaged over the frequency band with probability one. Numerical experiments are presented to illustrate the validity and effectiveness of the proposed method for the case that the random source is the white noise.
As Artificial Intelligence as a Service gains popularity, protecting well-trained models as intellectual property is becoming increasingly important. Generally speaking, there are two common protection methods: ownership verification and usage author
ization. In this paper, we propose Non-Transferable Learning (NTL), a novel approach that captures the exclusive data representation in the learned model and restricts the model generalization ability to certain domains. This approach provides effective solutions to both model verification and authorization. For ownership verification, watermarking techniques are commonly used but are often vulnerable to sophisticated watermark removal methods. Our NTL-based model verification approach instead provides robust resistance to state-of-the-art watermark removal methods, as shown in extensive experiments for four of such methods over the digits, CIFAR10 & STL10, and VisDA datasets. For usage authorization, prior solutions focus on authorizing specific users to use the model, but authorized users can still apply the model to any data without restriction. Our NTL-based authorization approach instead provides data-centric usage protection by significantly degrading the performance of usage on unauthorized data. Its effectiveness is also shown through experiments on a variety of datasets.
Strange quark matter, which is composed of u, d, and s quarks, could be the true ground of matter. According to this hypothesis, compact stars may actually be strange quark stars, and there may even be stable strange quark dwarfs and strange quark pl
anets. The detection of the binary neutron star merger event GW170817 provides us new clues on the equation of state of compact stars. In this study, the tidal deformability of strange quark planets and strange quark dwarfs are calculated. It is found that the tidal deformability of strange quark objects is smaller than that of normal matter counterparts. For a typical 0.6 M$_odot$ compact star, the tidal deformability of a strange dwarf is about 1.4 times less than that of a normal white dwarf. The difference is even more significant between strange quark planets and normal matter planets. Additionally, if the strange quark planet is a bare one (i.e., not covered by a normal matter curst), the tidal deformability will be extremely small, which means bare strange quark planets will hardly be distorted by tidal forces. Our study clearly proves the effectiveness of identifying strange quark objects via searching for strange quark planets through gravitational-wave observations.
Data augmentation is an approach that can effectively improve the performance of multimodal machine learning. This paper introduces a generative model for data augmentation by leveraging the correlations among multiple modalities. Different from conv
entional data augmentation approaches that apply low level operations with deterministic heuristics, our method proposes to learn an augmentation sampler that generates samples of the target modality conditioned on observed modalities in the variational auto-encoder framework. Additionally, the proposed model is able to quantify the confidence of augmented data by its generative probability, and can be jointly updated with a downstream pipeline. Experiments on Visual Question Answering tasks demonstrate the effectiveness of the proposed generative model, which is able to boost the strong UpDn-based models to the state-of-the-art performance.
Originally developed for imputing missing entries in low rank, or approximately low rank matrices, matrix completion has proven widely effective in many problems where there is no reason to assume low-dimensional linear structure in the underlying ma
trix, as would be imposed by rank constraints. In this manuscript, we build some theoretical intuition for this behavior. We consider matrices which are not necessarily low-rank, but lie in a low-dimensional non-linear manifold. We show that nuclear-norm penalization is still effective for recovering these matrices when observations are missing completely at random. In particular, we give upper bounds on the rate of convergence as a function of the number of rows, columns, and observed entries in the matrix, as well as the smoothness and dimension of the non-linear embedding. We additionally give a minimax lower bound: This lower bound agrees with our upper bound (up to a logarithmic factor), which shows that nuclear-norm penalization is (up to log terms) minimax rate optimal for these problems.
Previous studies have verified that the functionality of black-box models can be stolen with full probability outputs. However, under the more practical hard-label setting, we observe that existing methods suffer from catastrophic performance degrada
tion. We argue this is due to the lack of rich information in the probability prediction and the overfitting caused by hard labels. To this end, we propose a novel hard-label model stealing method termed emph{black-box dissector}, which consists of two erasing-based modules. One is a CAM-driven erasing strategy that is designed to increase the information capacity hidden in hard labels from the victim model. The other is a random-erasing-based self-knowledge distillation module that utilizes soft labels from the substitute model to mitigate overfitting. Extensive experiments on four widely-used datasets consistently demonstrate that our method outperforms state-of-the-art methods, with an improvement of at most $8.27%$. We also validate the effectiveness and practical potential of our method on real-world APIs and defense methods. Furthermore, our method promotes other downstream tasks, emph{i.e.}, transfer adversarial attacks.
Timely and reasonable matching of the control parameters and geological conditions of the rock mass in tunnel excavation is crucial for hard rock tunnel boring machines (TBMs). Therefore, this paper proposes an intelligent decision method for the mai
n control parameters of the TBM based on the multi-objective optimization of excavation efficiency and cost. The main objectives of this method are to obtain the most important parameters of the rock mass and machine, determine the optimization objective, and establish the objective function. In this study, muck information was included as an important parameter in the traditional rock mass and machine parameter database. The rock-machine interaction model was established through an improved neural network algorithm. Using 250 sets of data collected in the field, the validity of the rock-machine interaction relationship model was verified. Then, taking the cost as the optimization objective, the cost calculation model related to tunneling and the cutter was obtained. Subsequently, combined with rock-machine interaction model, the objective function of control parameter optimization based on cost was established. Finally, a tunneling test was carried out at the engineering site, and the main TBM control parameters (thrust and torque) after the optimization decision were used to excavate the test section. Compared with the values in the section where the TBM operators relied on experience, the average penetration rate of the TBM increased by 11.10%, and the average cutter life increased by 15.62%. The results indicate that this method can play an effective role in TBM tunneling in the test section.
