The domain of speech emotion recognition (SER) has persistently been a frontier within the landscape of machine learning. It is an active field that has been revolutionized in the last few decades and whose implementations are remarkable in multiple applications that could affect daily life. Consequently, the Iberian Languages Evaluation Forum (IberLEF) of 2024 held a competitive challenge to leverage the SER results with a Spanish corpus. This paper presents the approach followed with the goal of participating in this competition. The main architecture consists of different pre-trained speech and text models to extract features from both modalities, utilizing an attention pooling mechanism. The proposed system has achieved the first position in the challenge with an 86.69% in Macro F1-Score.
This paper investigates a movable antenna (MA)-assisted multiuser integrated sensing and communication (ISAC) system, where the base station (BS) and communication users are all equipped with MA for improving both the sensing and communication performance. We employ the Cramer-Rao bound (CRB) as the performance metric of sensing, thus a joint beamforming design and MAs' position optimizing problem is formulated to minimize the CRB. However the resulting optimization problem is NP-hard and the variables are highly coupled. To tackle this problem, we propose an alternating optimization (AO) framework by adopting semidefinite relaxation (SDR) and successive convex approximation (SCA) technique. Numerical results reveal that the proposed MA-assisted ISAC system achieves lower estimation CRB compared to the fixed-position antenna (FPA) counterpart.
The increasing volume of space objects in Earth's orbit presents a significant challenge for Space Situational Awareness (SSA). And in particular, accurate orbit prediction is crucial to anticipate the position and velocity of space objects, for collision avoidance and space debris mitigation. When performing Orbit Prediction (OP), it is necessary to consider the impact of non-conservative forces, such as atmospheric drag and gravitational perturbations, that contribute to uncertainty around the future position of spacecraft and space debris alike. Conventional propagator methods like the SGP4 inadequately account for these forces, while numerical propagators are able to model the forces at a high computational cost. To address these limitations, we propose an orbit prediction algorithm utilizing machine learning. This algorithm forecasts state vectors on a spacecraft using past positions and environmental variables like atmospheric density from external sources. The orbital data used in the paper is gathered from precision ephemeris data from the International Laser Ranging Service (ILRS), for the period of almost a year. We show how the use of machine learning and time-series techniques can produce low positioning errors at a very low computational cost, thus significantly improving SSA capabilities by providing faster and reliable orbit determination for an ever increasing number of space objects.
This study compares Term Frequency-Inverse Document Frequency (TF-IDF) features with Sentence Transformers for detecting writers' stances--favorable, opposing, or neutral--towards three significant topics: COVID-19 vaccine, digital transformation, and women empowerment. Through empirical evaluation, we demonstrate that Sentence Transformers outperform TF-IDF features across various experimental setups. Our team, dzStance, participated in a stance detection competition, achieving the 13th position (74.91%) among 15 teams in Women Empowerment, 10th (73.43%) in COVID Vaccine, and 12th (66.97%) in Digital Transformation. Overall, our team's performance ranked 13th (71.77%) among all participants. Notably, our approach achieved promising F1-scores, highlighting its effectiveness in identifying writers' stances on diverse topics. These results underscore the potential of Sentence Transformers to enhance stance detection models for addressing critical societal issues.
In this paper, a novel approach to sign language recognition based on state tying in each of data streams is presented. In this framework, it is assumed that hand gesture signal is represented in terms of six synchronous data streams, i.e., the left/right hand position, left/right hand orientation and left/right handshape. This approach offers a very accurate representation of the sign space and keeps the number of parameters reasonably small in favor of a fast decoding. Experiments were carried out for 5177 Chinese signs. The real time isolated recognition rate is 94.8%. For continuous sign recognition, the word correct rate is 91.4%. Keywords: Sign language recognition; Automatic sign language translation; Hand gesture recognition; Hidden Markov models; State-tying; Multimodal user interface; Virtual reality; Man-machine systems.
It has become common practice now to use random initialization schemes, rather than the pre-trained embeddings, when training transformer based models from scratch. Indeed, we find that pre-trained word embeddings from GloVe, and some sub-word embeddings extracted from language models such as T5 and mT5 fare much worse compared to random initialization. This is counter-intuitive given the well-known representational and transfer-learning advantages of pre-training. Interestingly, we also find that BERT and mBERT embeddings fare better than random initialization, showing the advantages of pre-trained representations. In this work, we posit two potential factors that contribute to these mixed results: the model sensitivity to parameter distribution and the embedding interactions with position encodings. We observe that pre-trained GloVe, T5, and mT5 embeddings have a wider distribution of values. As argued in the initialization studies, such large value initializations can lead to poor training because of saturated outputs. Further, the larger embedding values can, in effect, absorb the smaller position encoding values when added together, thus losing position information. Standardizing the pre-trained embeddings to a narrow range (e.g. as prescribed by Xavier) leads to substantial gains for Glove, T5, and mT5 embeddings. On the other hand, BERT pre-trained embeddings, while larger, are still relatively closer to Xavier initialization range which may allow it to effectively transfer the pre-trained knowledge.
This paper presents a general scheme for enhancing the convergence and performance of DETR (DEtection TRansformer). We investigate the slow convergence problem in transformers from a new perspective, suggesting that it arises from the self-attention that introduces no structural bias over inputs. To address this issue, we explore incorporating position relation prior as attention bias to augment object detection, following the verification of its statistical significance using a proposed quantitative macroscopic correlation (MC) metric. Our approach, termed Relation-DETR, introduces an encoder to construct position relation embeddings for progressive attention refinement, which further extends the traditional streaming pipeline of DETR into a contrastive relation pipeline to address the conflicts between non-duplicate predictions and positive supervision. Extensive experiments on both generic and task-specific datasets demonstrate the effectiveness of our approach. Under the same configurations, Relation-DETR achieves a significant improvement (+2.0% AP compared to DINO), state-of-the-art performance (51.7% AP for 1x and 52.1% AP for 2x settings), and a remarkably faster convergence speed (over 40% AP with only 2 training epochs) than existing DETR detectors on COCO val2017. Moreover, the proposed relation encoder serves as a universal plug-in-and-play component, bringing clear improvements for theoretically any DETR-like methods. Furthermore, we introduce a class-agnostic detection dataset, SA-Det-100k. The experimental results on the dataset illustrate that the proposed explicit position relation achieves a clear improvement of 1.3% AP, highlighting its potential towards universal object detection. The code and dataset are available at https://github.com/xiuqhou/Relation-DETR.
In this paper, we present DiT-MoE, a sparse version of the diffusion Transformer, that is scalable and competitive with dense networks while exhibiting highly optimized inference. The DiT-MoE includes two simple designs: shared expert routing and expert-level balance loss, thereby capturing common knowledge and reducing redundancy among the different routed experts. When applied to conditional image generation, a deep analysis of experts specialization gains some interesting observations: (i) Expert selection shows preference with spatial position and denoising time step, while insensitive with different class-conditional information; (ii) As the MoE layers go deeper, the selection of experts gradually shifts from specific spacial position to dispersion and balance. (iii) Expert specialization tends to be more concentrated at the early time step and then gradually uniform after half. We attribute it to the diffusion process that first models the low-frequency spatial information and then high-frequency complex information. Based on the above guidance, a series of DiT-MoE experimentally achieves performance on par with dense networks yet requires much less computational load during inference. More encouragingly, we demonstrate the potential of DiT-MoE with synthesized image data, scaling diffusion model at a 16.5B parameter that attains a new SoTA FID-50K score of 1.80 in 512$\times$512 resolution settings. The project page: https://github.com/feizc/DiT-MoE.
This short position paper provides a manually curated list of non-English image captioning datasets (as of May 2024). Through this list, we can observe the dearth of datasets in different languages: only 23 different languages are represented. With the addition of the Crossmodal-3600 dataset (Thapliyal et al., 2022, 36 languages) this number increases somewhat, but still this number is small compared to the +/-500 institutional languages that are out there. This paper closes with some open questions for the field of Vision & Language.
This article presents a three-dimensional nonlinear trajectory tracking control strategy for unmanned aerial vehicles (UAVs) in the presence of spatial constraints. As opposed to many existing control strategies, which do not consider spatial constraints, the proposed strategy considers spatial constraints on each degree of freedom movement of the UAV. Such consideration makes the design appealing for many practical applications, such as pipeline inspection, boundary tracking, etc. The proposed design accounts for the limited information about the inertia matrix, thereby affirming its inherent robustness against unmodeled dynamics and other imperfections. We rigorously show that the UAV will converge to its desired path by maintaining bounded position, orientation, and linear and angular speeds. Finally, we demonstrate the effectiveness of the proposed strategy through various numerical simulations.
In this paper, we investigate physical layer security (PLS) for full-duplex (FD) multi-user systems. We consider a base station (BS) that operates in FD mode and transmits artificial noise (AN) to simultaneously protect uplink (UL) and downlink (DL) transmissions. Conventional fixed-position antennas (FPAs) at the FD BS struggle to fully exploit spatial degrees of freedom (DoFs) to improve signal reception and suppress interference. To overcome this limitation, we propose a novel FD BS architecture equipped with multiple transmit and receive movable antennas (MAs). The MAs introduce the DoFs in antenna position optimization, which can improve the performance of secure communication systems. To serve users and counter the cooperative interception of multiple eavesdroppers (Eves), we formulate a sum of secrecy rates (SSR) maximization problem to jointly optimize the MA positions, the transmit, receive, and AN beamformers at the BS, and the UL powers. We propose an alternating optimization (AO) algorithm, which decomposes the original problem into three sub-problems, to solve the challenging non-convex optimization problem with highly coupled variables. Specifically, we propose the multi-velocity particle swarm optimization (MVPSO), which is an improved version of the standard particle swarm optimization (PSO), to simultaneously optimize all MA positions. The transmit/AN beamformers and the UL powers are solved by successive convex approximation (SCA). The optimal receive beamformer is derived as a closed-form solution. Simulation results demonstrate the effectiveness of the proposed algorithms and the advantages of MAs over conventional FPAs in enhancing the security of FD multi-user systems.
Micro-mobility transport vehicles such as e-scooters are joining current sidewalk users and affect the safety and comfort of pedestrians as primary sidewalk users. The lack of agreed-upon metrics to quantify people's discomfort hinders shared public space safety research. We introduce perceived Time To Collision (TTC) as a potential metric of user discomfort performing controlled experiments using an e-scooter and a pedestrian moving in a hallway. The results strongly correlate the participant's reported discomfort and the perceived TTC. Therefore, TTC is a potential metric for public space users' discomfort. Since the metric only uses relative velocity and position information, it is a viable candidate for neighboring people's discomfort estimation in advanced driver assistance systems for e-scooters and PMVs. Our ongoing research extends the results to mobile robots.
In three-dimensional geographical scenes, adding labels with leader lines to point features can significantly improve their visibility. Leadered labels have a large degree of freedom in position con-figuration, but existing methods are mostly based on limited position candidate models, which not only fail to effectively utilize the map space but also make it difficult to consider the relative relationships between labels. Therefore, we conceptualize the dynamic configuration process of computing label positions as akin to solving a map displacement problem. We use a triangulated graph to delineate spatial relationships among labels and calculate the forces exerted on labels considering the constraints associated with point feature labels. Then we use the Beams Displacement Method to iteratively calculate new positions for the labels. Our experimental outcomes demonstrate that this method effectively mitigates label overlay issues while maintaining minimal average directional deviation between adjacent labels. Furthermore, this method is adaptable to various types of leader line labels. Meanwhile, we also discuss the block processing strategy to improve the efficiency of label configuration and analyze the impact of different proximity graphs.
Learning rotation-invariant distinctive features is a fundamental requirement for point cloud registration. Existing methods often use rotation-sensitive networks to extract features, while employing rotation augmentation to learn an approximate invariant mapping rudely. This makes networks fragile to rotations, overweight, and hinders the distinctiveness of features. To tackle these problems, we propose a novel position-aware rotation-equivariant network, for efficient, light-weighted, and robust registration. The network can provide a strong model inductive bias to learn rotation-equivariant/invariant features, thus addressing the aforementioned limitations. To further improve the distinctiveness of descriptors, we propose a position-aware convolution, which can better learn spatial information of local structures. Moreover, we also propose a feature-based hypothesis proposer. It leverages rotation-equivariant features that encode fine-grained structure orientations to generate reliable model hypotheses. Each correspondence can generate a hypothesis, thus it is more efficient than classic estimators that require multiple reliable correspondences. Accordingly, a contrastive rotation loss is presented to enhance the robustness of rotation-equivariant features against data degradation. Extensive experiments on indoor and outdoor datasets demonstrate that our method significantly outperforms the SOTA methods in terms of registration recall while being lightweight and keeping a fast speed. Moreover, experiments on rotated datasets demonstrate its robustness against rotation variations. Code is available at https://github.com/yaorz97/PARENet.
Accurate depth information is crucial for enhancing the performance of multi-view 3D object detection. Despite the success of some existing multi-view 3D detectors utilizing pixel-wise depth supervision, they overlook two significant phenomena: 1) the depth supervision obtained from LiDAR points is usually distributed on the surface of the object, which is not so friendly to existing DETR-based 3D detectors due to the lack of the depth of 3D object center; 2) for distant objects, fine-grained depth estimation of the whole object is more challenging. Therefore, we argue that the object-wise depth (or 3D center of the object) is essential for accurate detection. In this paper, we propose a new multi-view 3D object detector named OPEN, whose main idea is to effectively inject object-wise depth information into the network through our proposed object-wise position embedding. Specifically, we first employ an object-wise depth encoder, which takes the pixel-wise depth map as a prior, to accurately estimate the object-wise depth. Then, we utilize the proposed object-wise position embedding to encode the object-wise depth information into the transformer decoder, thereby producing 3D object-aware features for final detection. Extensive experiments verify the effectiveness of our proposed method. Furthermore, OPEN achieves a new state-of-the-art performance with 64.4% NDS and 56.7% mAP on the nuScenes test benchmark.
This study addresses a critical gap in safety tuning practices for Large Language Models (LLMs) by identifying and tackling a refusal position bias within safety tuning data, which compromises the models' ability to appropriately refuse generating unsafe content. We introduce a novel approach, Decoupled Refusal Training (DeRTa), designed to empower LLMs to refuse compliance to harmful prompts at any response position, significantly enhancing their safety capabilities. DeRTa incorporates two novel components: (1) Maximum Likelihood Estimation (MLE) with Harmful Response Prefix, which trains models to recognize and avoid unsafe content by appending a segment of harmful response to the beginning of a safe response, and (2) Reinforced Transition Optimization (RTO), which equips models with the ability to transition from potential harm to safety refusal consistently throughout the harmful response sequence. Our empirical evaluation, conducted using LLaMA3 and Mistral model families across six attack scenarios, demonstrates that our method not only improves model safety without compromising performance but also surpasses well-known models such as GPT-4 in defending against attacks. Importantly, our approach successfully defends recent advanced attack methods (e.g., CodeAttack) that have jailbroken GPT-4 and LLaMA3-70B-Instruct. Our code and data can be found at https://github.com/RobustNLP/DeRTa.
This paper presents a novel quaternion-based nonsingular control system for underactuated vertical-take-off and landing (VTOL) Unmanned Aerial Vehicles (UAVs). Position and attitude tracking is challenging regarding singularity and accuracy. Quaternion-based Adaptive Backstepping Control (QABC) is developed to tackle the underactuated issues of UAV control systems in a cascaded way. Leveraging the virtual control (auxiliary control) developed in the QABC, desired attitude components and required thrust are produced. Afterwards, we propose Quaternion-based Sliding Mode Control (QASMC) to enhance the stability and mitigate chattering issues. The sliding surface is modified to avoid singularity compared to conventional SMC. To improve the robustness of controllers, the control parameters are updated using adaptation laws. Furthermore, the asymptotic stability of translational and rotational dynamics is guaranteed by utilizing Lyapunov stability and Barbalet Lemma. Finally, the comprehensive comparison results are provided to verify the effectiveness of the proposed controllers in the presence of unknown time-varying parameter uncertainties and significant initial errors. Keywords: Non-singular Sliding Mode Control, Adaptive Backstepping Control, Unit-quaternion, Drones, Unmanned Aerial Vehicles, Asymptotic Stability, Position and Orientation Control
Machine learning (ML) datasets, often perceived as neutral, inherently encapsulate abstract and disputed social constructs. Dataset curators frequently employ value-laden terms such as diversity, bias, and quality to characterize datasets. Despite their prevalence, these terms lack clear definitions and validation. Our research explores the implications of this issue by analyzing "diversity" across 135 image and text datasets. Drawing from social sciences, we apply principles from measurement theory to identify considerations and offer recommendations for conceptualizing, operationalizing, and evaluating diversity in datasets. Our findings have broader implications for ML research, advocating for a more nuanced and precise approach to handling value-laden properties in dataset construction.
Visual Place Recognition (VPR) systems often have imperfect performance, affecting the `integrity' of position estimates and subsequent robot navigation decisions. Previously, SVM classifiers have been used to monitor VPR integrity. This research introduces a novel Multi-Layer Perceptron (MLP) integrity monitor which demonstrates improved performance and generalizability, removing per-environment training and reducing manual tuning requirements. We test our proposed system in extensive real-world experiments, presenting two real-time integrity-based VPR verification methods: a single-query rejection method for robot navigation to a goal zone (Experiment 1); and a history-of-queries method that takes a best, verified, match from its recent trajectory and uses an odometer to extrapolate a current position estimate (Experiment 2). Noteworthy results for Experiment 1 include a decrease in aggregate mean along-track goal error from ~9.8m to ~3.1m, and an increase in the aggregate rate of successful mission completion from ~41% to ~55%. Experiment 2 showed a decrease in aggregate mean along-track localization error from ~2.0m to ~0.5m, and an increase in the aggregate localization precision from ~97% to ~99%. Overall, our results demonstrate the practical usefulness of a VPR integrity monitor in real-world robotics to improve VPR localization and consequent navigation performance.
High-resolution inputs enable Large Vision-Language Models (LVLMs) to discern finer visual details, enhancing their comprehension capabilities. To reduce the training and computation costs caused by high-resolution input, one promising direction is to use sliding windows to slice the input into uniform patches, each matching the input size of the well-trained vision encoder. Although efficient, this slicing strategy leads to the fragmentation of original input, i.e., the continuity of contextual information and spatial geometry is lost across patches, adversely affecting performance in cross-patch context perception and position-specific tasks. To overcome these shortcomings, we introduce HiRes-LLaVA, a novel framework designed to efficiently process any size of high-resolution input without altering the original contextual and geometric information. HiRes-LLaVA comprises two innovative components: (i) a SliceRestore adapter that reconstructs sliced patches into their original form, efficiently extracting both global and local features via down-up-sampling and convolution layers, and (ii) a Self-Mining Sampler to compresses the vision tokens based on themselves, preserving the original context and positional information while reducing training overhead. To assess the ability of handling context fragmentation, we construct a new benchmark, EntityGrid-QA, consisting of edge-related and position-related tasks. Our comprehensive experiments demonstrate the superiority of HiRes-LLaVA on both existing public benchmarks and on EntityGrid-QA, particularly on document-oriented tasks, establishing new standards for handling high-resolution inputs.