PyTorchKR Papers

최신 논문

총 715편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2111.12449공식 코드 · Papers with Code

Background-Click Supervision for Temporal Action Localization

Weakly supervised temporal action localization aims at learning the instance-level action pattern from the video-level labels, where a significant challenge is action-context confusion. To overcome this challenge, one recent work builds an action-click supervision framework. It requires similar annotation costs but can steadily improve the localization performance when compared to the conventional weakly supervised methods. In this paper, by revealing that the performance bottleneck of the existing approaches mainly comes from the background errors, we find that a stronger action localizer can be trained with labels on the background video frames rather than those on the action frames. To this end, we convert the action-click supervision to the background-click supervision and develop a novel method, called BackTAL. Specifically, BackTAL implements two-fold modeling on the background video frames, i.e. the position modeling and the feature modeling. In position modeling, we not only conduct supervised learning on the annotated video frames but also design a score separation module to enlarge the score differences between the potential action frames and backgrounds. In feature modeling, we propose an affinity module to measure frame-specific similarities among neighboring frames and dynamically attend to informative neighbors when calculating temporal convolution. Extensive experiments on three benchmarks are conducted, which demonstrate the high performance of the established BackTAL and the rationality of the proposed background-click supervision. Code is available at https://github.com/VividLe/BackTAL.

arXiv:2111.11011공식 코드 · Papers with Code

CDistNet: Perceiving Multi-Domain Character Distance for Robust Text Recognition

The Transformer-based encoder-decoder framework is becoming popular in scene text recognition, largely because it naturally integrates recognition clues from both visual and semantic domains. However, recent studies show that the two kinds of clues are not always well registered and therefore, feature and character might be misaligned in difficult text (e.g., with a rare shape). As a result, constraints such as character position are introduced to alleviate this problem. Despite certain success, visual and semantic are still separately modeled and they are merely loosely associated. In this paper, we propose a novel module called Multi-Domain Character Distance Perception (MDCDP) to establish a visually and semantically related position embedding. MDCDP uses the position embedding to query both visual and semantic features following the cross-attention mechanism. The two kinds of clues are fused into the position branch, generating a content-aware embedding that well perceives character spacing and orientation variants, character semantic affinities, and clues tying the two kinds of information. They are summarized as the multi-domain character distance. We develop CDistNet that stacks multiple MDCDPs to guide a gradually precise distance modeling. Thus, the feature-character alignment is well built even various recognition difficulties are presented. We verify CDistNet on ten challenging public datasets and two series of augmented datasets created by ourselves. The experiments demonstrate that CDistNet performs highly competitively. It not only ranks top-tier in standard benchmarks, but also outperforms recent popular methods by obvious margins on real and augmented datasets presenting severe text deformation, poor linguistic support, and rare character layouts. Code is available at https://github.com/simplify23/CDistNet.

arXiv:2111.08108공식 코드 · Papers with Code

Physics-informed neural networks via stochastic Hamiltonian dynamics learning

In this paper, we propose novel learning frameworks to tackle optimal control problems by applying the Pontryagin maximum principle and then solving for a Hamiltonian dynamical system. Applying the Pontryagin maximum principle to the original optimal control problem shifts the learning focus to reduced Hamiltonian dynamics and corresponding adjoint variables. Then, the reduced Hamiltonian networks can be learned by going backwards in time and then minimizing loss function deduced from the Pontryagin maximum principle's conditions. The learning process is further improved by progressively learning a posterior distribution of the reduced Hamiltonians. This is achieved through utilizing a variational autoencoder which leads to more effective path exploration process. We apply our learning frameworks called NeuralPMP to various control tasks and obtain competitive results.

arXiv:2111.07574공식 코드 · Papers with Code

Vision-Position Multi-Modal Beam Prediction Using Real Millimeter Wave Datasets

Enabling highly-mobile millimeter wave (mmWave) and terahertz (THz) wireless communication applications requires overcoming the critical challenges associated with the large antenna arrays deployed at these systems. In particular, adjusting the narrow beams of these antenna arrays typically incurs high beam training overhead that scales with the number of antennas. To address these challenges, this paper proposes a multi-modal machine learning based approach that leverages positional and visual (camera) data collected from the wireless communication environment for fast beam prediction. The developed framework has been tested on a real-world vehicular dataset comprising practical GPS, camera, and mmWave beam training data. The results show the proposed approach achieves more than $\approx$ 75\% top-1 beam prediction accuracy and close to 100\% top-3 beam prediction accuracy in realistic communication scenarios.

arXiv:2111.06643공식 코드 · Papers with Code

Fully Automatic Page Turning on Real Scores

We present a prototype of an automatic page turning system that works directly on real scores, i.e., sheet images, without any symbolic representation. Our system is based on a multi-modal neural network architecture that observes a complete sheet image page as input, listens to an incoming musical performance, and predicts the corresponding position in the image. Using the position estimation of our system, we use a simple heuristic to trigger a page turning event once a certain location within the sheet image is reached. As a proof of concept we further combine our system with an actual machine that will physically turn the page on command.

arXiv:2111.06599공식 코드 · Papers with Code

PESTO: Switching Point based Dynamic and Relative Positional Encoding for Code-Mixed Languages

NLP applications for code-mixed (CM) or mix-lingual text have gained a significant momentum recently, the main reason being the prevalence of language mixing in social media communications in multi-lingual societies like India, Mexico, Europe, parts of USA etc. Word embeddings are basic build-ing blocks of any NLP system today, yet, word embedding for CM languages is an unexplored territory. The major bottleneck for CM word embeddings is switching points, where the language switches. These locations lack in contextually and statistical systems fail to model this phenomena due to high variance in the seen examples. In this paper we present our initial observations on applying switching point based positional encoding techniques for CM language, specifically Hinglish (Hindi - English). Results are only marginally better than SOTA, but it is evident that positional encoding could bean effective way to train position sensitive language models for CM text.

arXiv:2111.05100공식 코드 · Papers with Code

EEGEyeNet: a Simultaneous Electroencephalography and Eye-tracking Dataset and Benchmark for Eye Movement Prediction

We present a new dataset and benchmark with the goal of advancing research in the intersection of brain activities and eye movements. Our dataset, EEGEyeNet, consists of simultaneous Electroencephalography (EEG) and Eye-tracking (ET) recordings from 356 different subjects collected from three different experimental paradigms. Using this dataset, we also propose a benchmark to evaluate gaze prediction from EEG measurements. The benchmark consists of three tasks with an increasing level of difficulty: left-right, angle-amplitude and absolute position. We run extensive experiments on this benchmark in order to provide solid baselines, both based on classical machine learning models and on large neural networks. We release our complete code and data and provide a simple and easy-to-use interface to evaluate new methods.

arXiv:2110.12486공식 코드 · Papers with Code

EgoNN: Egocentric Neural Network for Point Cloud Based 6DoF Relocalization at the City Scale

The paper presents a deep neural network-based method for global and local descriptors extraction from a point cloud acquired by a rotating 3D LiDAR. The descriptors can be used for two-stage 6DoF relocalization. First, a course position is retrieved by finding candidates with the closest global descriptor in the database of geo-tagged point clouds. Then, the 6DoF pose between a query point cloud and a database point cloud is estimated by matching local descriptors and using a robust estimator such as RANSAC. Our method has a simple, fully convolutional architecture based on a sparse voxelized representation. It can efficiently extract a global descriptor and a set of keypoints with local descriptors from large point clouds with tens of thousand points. Our code and pretrained models are publicly available on the project website.

arXiv:2110.05158공식 코드 · Papers with Code

Can the brain use waves to solve planning problems?

A variety of behaviors like spatial navigation or bodily motion can be formulated as graph traversal problems through cognitive maps. We present a neural network model which can solve such tasks and is compatible with a broad range of empirical findings about the mammalian neocortex and hippocampus. The neurons and synaptic connections in the model represent structures that can result from self-organization into a cognitive map via Hebbian learning, i.e. into a graph in which each neuron represents a point of some abstract task-relevant manifold and the recurrent connections encode a distance metric on the manifold. Graph traversal problems are solved by wave-like activation patterns which travel through the recurrent network and guide a localized peak of activity onto a path from some starting position to a target state.

arXiv:2110.00876공식 코드 · Papers with Code

Incremental Non-Gaussian Inference for SLAM Using Normalizing Flows

This paper presents normalizing flows for incremental smoothing and mapping (NF-iSAM), a novel algorithm for inferring the full posterior distribution in SLAM problems with nonlinear measurement models and non-Gaussian factors. NF-iSAM exploits the expressive power of neural networks, and trains normalizing flows to model and sample the full posterior. By leveraging the Bayes tree, NF-iSAM enables efficient incremental updates similar to iSAM2, albeit in the more challenging non-Gaussian setting. We demonstrate the advantages of NF-iSAM over state-of-the-art point and distribution estimation algorithms using range-only SLAM problems with data association ambiguity. NF-iSAM presents superior accuracy in describing the posterior beliefs of continuous variables (e.g., position) and discrete variables (e.g., data association).

arXiv:2110.00335공식 코드 · Papers with Code

Geometry Attention Transformer with Position-aware LSTMs for Image Captioning

In recent years, transformer structures have been widely applied in image captioning with impressive performance. For good captioning results, the geometry and position relations of different visual objects are often thought of as crucial information. Aiming to further promote image captioning by transformers, this paper proposes an improved Geometry Attention Transformer (GAT) model. In order to further leverage geometric information, two novel geometry-aware architectures are designed respectively for the encoder and decoder in our GAT. Besides, this model includes the two work modules: 1) a geometry gate-controlled self-attention refiner, for explicitly incorporating relative spatial information into image region representations in encoding steps, and 2) a group of position-LSTMs, for precisely informing the decoder of relative word position in generating caption texts. The experiment comparisons on the datasets MS COCO and Flickr30K show that our GAT is efficient, and it could often outperform current state-of-the-art image captioning models.

arXiv:2109.15308공식 코드 · Papers with Code

MOLUCINATE: A Generative Model for Molecules in 3D Space

Recent advances in machine learning have enabled generative models for both optimization and de novo generation of drug candidates with desired properties. Previous generative models have focused on producing SMILES strings or 2D molecular graphs, while attempts at producing molecules in 3D have focused on reinforcement learning (RL), distance matrices, and pure atom density grids. Here we present MOLUCINATE (MOLecUlar ConvolutIoNal generATive modEl), a novel architecture that simultaneously generates topological and 3D atom position information. We demonstrate the utility of this method by using it to optimize molecules for desired radius of gyration. In the future, this model can be used for more useful optimization such as binding affinity for a protein target.

arXiv:2109.11749공식 코드 · Papers with Code

Fine-Grained Image Generation from Bangla Text Description using Attentional Generative Adversarial Network

Generating fine-grained, realistic images from text has many applications in the visual and semantic realm. Considering that, we propose Bangla Attentional Generative Adversarial Network (AttnGAN) that allows intensified, multi-stage processing for high-resolution Bangla text-to-image generation. Our model can integrate the most specific details at different sub-regions of the image. We distinctively concentrate on the relevant words in the natural language description. This framework has achieved a better inception score on the CUB dataset. For the first time, a fine-grained image is generated from Bangla text using attentional GAN. Bangla has achieved 7th position among 100 most spoken languages. This inspires us to explicitly focus on this language, which will ensure the inevitable need of many people. Moreover, Bangla has a more complex syntactic structure and less natural language processing resource that validates our work more.

arXiv:2109.10473공식 코드 · Papers with Code

MVM3Det: A Novel Method for Multi-view Monocular 3D Detection

Monocular 3D object detection encounters occlusion problems in many application scenarios, such as traffic monitoring, pedestrian monitoring, etc., which leads to serious false negative. Multi-view object detection effectively solves this problem by combining data from different perspectives. However, due to label confusion and feature confusion, the orientation estimation of multi-view 3D object detection is intractable, which is important for object tracking and intention prediction. In this paper, we propose a novel multi-view 3D object detection method named MVM3Det which simultaneously estimates the 3D position and orientation of the object according to the multi-view monocular information. The method consists of two parts: 1) Position proposal network, which integrates the features from different perspectives into consistent global features through feature orthogonal transformation to estimate the position. 2) Multi-branch orientation estimation network, which introduces feature perspective pooling to overcome the two confusion problems during the orientation estimation. In addition, we present a first dataset for multi-view 3D object detection named MVM3D. Comparing with State-Of-The-Art (SOTA) methods on our dataset and public dataset WildTrack, our method achieves very competitive results.

arXiv:2109.09975공식 코드 · Papers with Code

Fast nonlinear risk assessment for autonomous vehicles using learned conditional probabilistic models of agent futures

This paper presents fast non-sampling based methods to assess the risk for trajectories of autonomous vehicles when probabilistic predictions of other agents' futures are generated by deep neural networks (DNNs). The presented methods address a wide range of representations for uncertain predictions including both Gaussian and non-Gaussian mixture models to predict both agent positions and control inputs conditioned on the scene contexts. We show that the problem of risk assessment when Gaussian mixture models (GMMs) of agent positions are learned can be solved rapidly to arbitrary levels of accuracy with existing numerical methods. To address the problem of risk assessment for non-Gaussian mixture models of agent position, we propose finding upper bounds on risk using nonlinear Chebyshev's Inequality and sums-of-squares (SOS) programming; they are both of interest as the former is much faster while the latter can be arbitrarily tight. These approaches only require higher order statistical moments of agent positions to determine upper bounds on risk. To perform risk assessment when models are learned for agent control inputs as opposed to positions, we propagate the moments of uncertain control inputs through the nonlinear motion dynamics to obtain the exact moments of uncertain position over the planning horizon. To this end, we construct deterministic linear dynamical systems that govern the exact time evolution of the moments of uncertain position in the presence of uncertain control inputs. The presented methods are demonstrated on realistic predictions from DNNs trained on the Argoverse and CARLA datasets and are shown to be effective for rapidly assessing the probability of low probability events.

arXiv:2109.09533공식 코드 · Papers with Code

Modeling Annotation Uncertainty with Gaussian Heatmaps in Landmark Localization

In landmark localization, due to ambiguities in defining their exact position, landmark annotations may suffer from large observer variabilities, which result in uncertain annotations. To model the annotation ambiguities of the training dataset, we propose to learn anisotropic Gaussian parameters modeling the shape of the target heatmap during optimization. Furthermore, our method models the prediction uncertainty of individual samples by fitting anisotropic Gaussian functions to the predicted heatmaps during inference. Besides state-of-the-art results, our experiments on datasets of hand radiographs and lateral cephalograms also show that Gaussian functions are correlated with both localization accuracy and observer variability. As a final experiment, we show the importance of integrating the uncertainty into decision making by measuring the influence of the predicted location uncertainty on the classification of anatomical abnormalities in lateral cephalograms.

arXiv:2109.05797공식 코드 · Papers with Code

Show Me How To Revise: Improving Lexically Constrained Sentence Generation with XLNet

Lexically constrained sentence generation allows the incorporation of prior knowledge such as lexical constraints into the output. This technique has been applied to machine translation, and dialog response generation. Previous work usually used Markov Chain Monte Carlo (MCMC) sampling to generate lexically constrained sentences, but they randomly determined the position to be edited and the action to be taken, resulting in many invalid refinements. To overcome this challenge, we used a classifier to instruct the MCMC-based models where and how to refine the candidate sentences. First, we developed two methods to create synthetic data on which the pre-trained model is fine-tuned to obtain a reliable classifier. Next, we proposed a two-step approach, "Predict and Revise", for constrained sentence generation. During the predict step, we leveraged the classifier to compute the learned prior for the candidate sentence. During the revise step, we resorted to MCMC sampling to revise the candidate sentence by conducting a sampled action at a sampled position drawn from the learned prior. We compared our proposed models with many strong baselines on two tasks, generating sentences with lexical constraints and text infilling. Experimental results have demonstrated that our proposed model performs much better than the previous work in terms of sentence fluency and diversity. Our code and pre-trained models are available at https://github.com/NLPCode/MCMCXLNet.

arXiv:2109.05644공식 코드 · Papers with Code

SHAPE: Shifted Absolute Position Embedding for Transformers

Position representation is crucial for building position-aware representations in Transformers. Existing position representations suffer from a lack of generalization to test data with unseen lengths or high computational cost. We investigate shifted absolute position embedding (SHAPE) to address both issues. The basic idea of SHAPE is to achieve shift invariance, which is a key property of recent successful position representations, by randomly shifting absolute positions during training. We demonstrate that SHAPE is empirically comparable to its counterpart while being simpler and faster.

arXiv:2109.04861공식 코드 · Papers with Code

GPS-Denied Navigation Using Low-Cost Inertial Sensors and Recurrent Neural Networks

Autonomous missions of drones require continuous and reliable estimates for the drone's attitude, velocity, and position. Traditionally, these states are estimated by applying Extended Kalman Filter (EKF) to Accelerometer, Gyroscope, Barometer, Magnetometer, and GPS measurements. When the GPS signal is lost, position and velocity estimates deteriorate quickly, especially when using low-cost inertial sensors. This paper proposes an estimation method that uses a Recurrent Neural Network (RNN) to allow reliable estimation of a drone's position and velocity in the absence of GPS signal. The RNN is trained on a public dataset collected using Pixhawk. This low-cost commercial autopilot logs the raw sensor measurements (network inputs) and corresponding EKF estimates (ground truth outputs). The dataset is comprised of 548 different flight logs with flight durations ranging from 4 to 32 minutes. For training, 465 flights are used, totaling 45 hours. The remaining 83 flights totaling 8 hours are held out for validation. Error in a single flight is taken to be the maximum absolute difference in 3D position (MPE) between the RNN predictions (without GPS) and the ground truth (EKF with GPS). On the validation set, the median MPE is 35 meters. MPE values as low as 2.7 meters in a 5-minutes flight could be achieved using the proposed method. The MPE in 90% of the validation flights is bounded below 166 meters. The network was experimentally tested and worked in real-time.

arXiv:2109.02377공식 코드 · Papers with Code

PermuteFormer: Efficient Relative Position Encoding for Long Sequences

A recent variation of Transformer, Performer, scales Transformer to longer sequences with a linear attention mechanism. However, it is not compatible with relative position encoding, which has advantages over absolute position encoding. In this paper, we discuss possible ways to add relative position encoding to Performer. Based on the analysis, we propose PermuteFormer, a Performer-based model with relative position encoding that scales linearly on long sequences. PermuteFormer applies position-dependent transformation on queries and keys to encode positional information into the attention module. This transformation is carefully crafted so that the final output of self-attention is not affected by absolute positions of tokens. PermuteFormer introduces negligible computational overhead by design that it runs as fast as Performer. We evaluate PermuteFormer on Long-Range Arena, a dataset for long sequences, as well as WikiText-103, a language modeling dataset. The experiments show that PermuteFormer uniformly improves the performance of Performer with almost no computational overhead and outperforms vanilla Transformer on most of the tasks.