Searcharxiv⌕ Search

arXiv subjects

Shilong Zhang

Publications and source records attributed to Shilong Zhang.

46 records · Page 3Linked to original sources

What Are Expected Queries in End-to-End Object Detection?

End-to-end object detection is rapidly progressed after the emergence of DETR. DETRs use a set of sparse queries that replace the dense candidate boxes in most traditional detectors. In comparison, the sparse queries cannot guarantee a high recall as dense priors. However, making queries dense is not trivial in current frameworks. It not only suffers from heavy computational cost but also difficult optimization. As both sparse and dense queries are imperfect, then \emph{what are expected queries in end-to-end object detection}? This paper shows that the expected queries should be Dense Distinct Queries (DDQ). Concretely, we introduce dense priors back to the framework to generate dense queries. A duplicate query removal pre-process is applied to these queries so that they are distinguishable from each other. The dense distinct queries are then iteratively processed to obtain final sparse outputs. We show that DDQ is stronger, more robust, and converges faster. It obtains 44.5 AP on the MS COCO detection dataset with only 12 epochs. DDQ is also robust as it outperforms previous methods on both object detection and instance segmentation tasks on various datasets. DDQ blends advantages from traditional dense priors and recent end-to-end detectors. We hope it can serve as a new baseline and inspires researchers to revisit the complementarity between traditional methods and end-to-end detectors. The source code is publicly available at \url{https://github.com/jshilong/DDQ}.

cs.CV↗

Group R-CNN for Weakly Semi-supervised Object Detection with Points

We study the problem of weakly semi-supervised object detection with points (WSSOD-P), where the training data is combined by a small set of fully annotated images with bounding boxes and a large set of weakly-labeled images with only a single point annotated for each instance. The core of this task is to train a point-to-box regressor on well-labeled images that can be used to predict credible bounding boxes for each point annotation. We challenge the prior belief that existing CNN-based detectors are not compatible with this task. Based on the classic R-CNN architecture, we propose an effective point-to-box regressor: Group R-CNN. Group R-CNN first uses instance-level proposal grouping to generate a group of proposals for each point annotation and thus can obtain a high recall rate. To better distinguish different instances and improve precision, we propose instance-level proposal assignment to replace the vanilla assignment strategy adopted in the original R-CNN methods. As naive instance-level assignment brings converging difficulty, we propose instance-aware representation learning which consists of instance-aware feature enhancement and instance-aware parameter generation to overcome this issue. Comprehensive experiments on the MS-COCO benchmark demonstrate the effectiveness of our method. Specifically, Group R-CNN significantly outperforms the prior method Point DETR by 3.9 mAP with 5% well-labeled images, which is the most challenging scenario. The source code can be found at https://github.com/jshilong/GroupRCNN

cs.CV↗

Rota-Baxter Lie $2$-algebras

In this paper, we introduce the notion of Rota-Baxter Lie $2$-algebras, which is a categorification of Rota-Baxter Lie algebras. We prove that the category of Rota-Baxter Lie $2$-algebras and the category of $2$-term Rota-Baxter $L_\infty$-algebras are equivalent. We introduce the notion of a crossed module of Rota-Baxter Lie algebras and show that there is a one-to-one correspondence between strict $2$-term Rota-Baxter $L_\infty$-algebras and crossed modules of Rota-Baxter Lie algebras. We give the construction of crossed modules of Lie algebras from crossed modules of Rota-Baxter Lie algebras.

math.CT↗

Dispersionless orbital excitations in (Li,Fe)OHFeSe superconductors

The superconducting critical temperature $T_{\mathrm{c}}$ of intercalated iron-selenide superconductor (Li,Fe)OHFeSe (FeSe11111) can be increased to 42 K from 8 K of bulk FeSe. It shows remarkably similar electronic properties as the high-$T_{\mathrm{c}}$ monolayer FeSe and provides a bulk counterpart to investigate the origin of enhanced superconductivity. Unraveling the nature of excitations is crucial for understanding the pairing mechanism in high-$T_{\mathrm{c}}$ iron selenides. Here we use resonant inelastic x-ray scattering (RIXS) to investigate the excitations in FeSe11111. Our high-quality data exhibit several Raman-like excitations, which are dispersionless and isotropic in momentum transfer and robust against varying $T_{\mathrm{c}}$. Using atomic multiplet calculations, we assign the low-energy $\sim 0.3$ and 0.7 eV Raman peaks as local $e_g-e_g$ and $e_g-t_{2g}$ orbital excitations. The intensity of these two features decreases with increasing temperature, suggesting a primary contribution of the orbital fluctuations. Our results highlight the importance of orbital degree of freedom for high-$T_{\mathrm{c}}$ iron selenides.

cond-mat.supr-con↗

Cohomologies and deformations of pre-Lie-morphism triples

A (pre-)Lie-morphism triple consists of two (pre-)Lie algebras and a (pre-)Lie algebra homomorphism between them. We give chomologies of pre-Lie-morphism triples. As an application, we study the infinitesimal deformations of pre-Lie-morphism triples. Finally, we show that the cohomology of the pre-Lie-morphism triple can be deduced from a new cohomology of the Lie-morphism triple.

math.RA↗

Group Fisher Pruning for Practical Network Compression

Network compression has been widely studied since it is able to reduce the memory and computation cost during inference. However, previous methods seldom deal with complicated structures like residual connections, group/depth-wise convolution and feature pyramid network, where channels of multiple layers are coupled and need to be pruned simultaneously. In this paper, we present a general channel pruning approach that can be applied to various complicated structures. Particularly, we propose a layer grouping algorithm to find coupled channels automatically. Then we derive a unified metric based on Fisher information to evaluate the importance of a single channel and coupled channels. Moreover, we find that inference speedup on GPUs is more correlated with the reduction of memory rather than FLOPs, and thus we employ the memory reduction of each channel to normalize the importance. Our method can be used to prune any structures including those with coupled channels. We conduct extensive experiments on various backbones, including the classic ResNet and ResNeXt, mobile-friendly MobileNetV2, and the NAS-based RegNet, both on image classification and object detection which is under-explored. Experimental results validate that our method can effectively prune sophisticated networks, boosting inference speed without sacrificing accuracy.

cs.CV↗

Scale-Equalizing Pyramid Convolution for Object Detection

Feature pyramid has been an efficient method to extract features at different scales. Development over this method mainly focuses on aggregating contextual information at different levels while seldom touching the inter-level correlation in the feature pyramid. Early computer vision methods extracted scale-invariant features by locating the feature extrema in both spatial and scale dimension. Inspired by this, a convolution across the pyramid level is proposed in this study, which is termed pyramid convolution and is a modified 3-D convolution. Stacked pyramid convolutions directly extract 3-D (scale and spatial) features and outperforms other meticulously designed feature fusion modules. Based on the viewpoint of 3-D convolution, an integrated batch normalization that collects statistics from the whole feature pyramid is naturally inserted after the pyramid convolution. Furthermore, we also show that the naive pyramid convolution, together with the design of RetinaNet head, actually best applies for extracting features from a Gaussian pyramid, whose properties can hardly be satisfied by a feature pyramid. In order to alleviate this discrepancy, we build a scale-equalizing pyramid convolution (SEPC) that aligns the shared pyramid convolution kernel only at high-level feature maps. Being computationally efficient and compatible with the head design of most single-stage object detectors, the SEPC module brings significant performance improvement ($>4$AP increase on MS-COCO2017 dataset) in state-of-the-art one-stage object detectors, and a light version of SEPC also has $\sim3.5$AP gain with only around 7% inference time increase. The pyramid convolution also functions well as a stand-alone module in two-stage object detectors and is able to improve the performance by $\sim2$AP. The source code can be found at https://github.com/jshilong/SEPC.

cs.CV↗

Extensions of Operators, Liftings of Monads and Distributive Laws

In a previous study, the algebraic formulation of the First Fundamental Theorem of Calculus (FFTC) is shown to allow extensions of differential and Rota-Baxter operators on the one hand, and to give rise to categorical explanations using the ideas of liftings of monads and comonads, and mixed distributive laws on the other. Generalizing the FFTC, we consider in this paper a class of constraints between a differential operator and a Rota-Baxter operator. For a given constraint, we show that the existences of extensions of differential and Rota-Baxter operators, of liftings of monads and comonads, and of mixed distributive laws are equivalent.

math.RA↗

Classification of operator extensions, monad liftings and distributive laws for differential algebras and Rota-Baxter algebras

Generalizing the algebraic formulation of the First Fundamental Theorem of Calculus (FFTC), a class of constraints involving a pair of operators was considered in \cite{ZGK2}. For a given constraint, the existences of extensions of differential and Rota-Baxter operators, of liftings of monads and comonads, and of mixed distributive laws are shown to be equivalent. In this paper, we give a classification of the constraints satisfying these equivalent conditions.

math.AC↗

Monads and distributive laws for Rota-Baxter and differential algebras

In recent years, algebraic studies of the differential calculus and integral calculus in the forms of differential algebra and Rota-Baxter algebra have been merged together to reflect the close relationship between the two calculi through the First Fundamental Theorem of Calculus. In this paper we study this relationship from a categorical point of view in the context of distributive laws which can be tracked back to the distributive law of multiplication over addition. The monad giving Rota-Baxter algebras and the comonad giving differential algebras are constructed. Then a mixed distributive law of the monad over the comonad is established. As a consequence, we obtain monads and comonads giving the composite structures of differential and Rota-Baxter algebras.

math.CT↗