Tao Ma

Xin Li, Tao Ma, Yuenan Hou et al. · stanford

LiDAR-camera fusion methods have shown impressive performance in 3D object detection. Recent advanced multi-modal methods mainly perform global fusion, where image features and point cloud features are fused across the whole scene. Such practice lacks fine-grained region-level information, yielding suboptimal fusion performance. In this paper, we present the novel Local-to-Global fusion network (LoGoNet), which performs LiDAR-camera fusion at both local and global levels. Concretely, the Global Fusion (GoF) of LoGoNet is built upon previous literature, while we exclusively use point centroids to more precisely represent the position of voxel features, thus achieving better cross-modal alignment. As to the Local Fusion (LoF), we first divide each proposal into uniform grids and then project these grid centers to the images. The image features around the projected grid points are sampled to be fused with position-decorated point cloud features, maximally utilizing the rich contextual information around the proposals. The Feature Dynamic Aggregation (FDA) module is further proposed to achieve information interaction between these locally and globally fused features, thus producing more informative multi-modal features. Extensive experiments on both Waymo Open Dataset (WOD) and KITTI datasets show that LoGoNet outperforms all state-of-the-art 3D detection methods. Notably, LoGoNet ranks 1st on Waymo 3D object detection leaderboard and obtains 81.02 mAPH (L2) detection performance. It is noteworthy that, for the first time, the detection performance on three classes surpasses 80 APH (L2) simultaneously. Code will be available at \url{https://github.com/sankin97/LoGoNet}.

Guohang Yan, Liu Zhuochun, Chengjie Wang et al. · stanford

Accurate sensor calibration is a prerequisite for multi-sensor perception and localization systems for autonomous vehicles. The intrinsic parameter calibration of the sensor is to obtain the mapping relationship inside the sensor, and the extrinsic parameter calibration is to transform two or more sensors into a unified spatial coordinate system. Most sensors need to be calibrated after installation to ensure the accuracy of sensor measurements. To this end, we present OpenCalib, a calibration toolbox that contains a rich set of various sensor calibration methods. OpenCalib covers manual calibration tools, automatic calibration tools, factory calibration tools, and online calibration tools for different application scenarios. At the same time, to evaluate the calibration accuracy and subsequently improve the accuracy of the calibration algorithm, we released a corresponding benchmark dataset. This paper introduces various features and calibration methods of this toolbox. To our knowledge, this is the first open-sourced calibration codebase containing the full set of autonomous-driving-related calibration approaches in this area. We wish that the toolbox could be helpful to autonomous driving researchers. We have open-sourced our code on GitHub to benefit the community. Code is available at https://github.com/PJLab-ADG/SensorsCalibration.

Licheng Wen, Daocheng Fu, Xin Li et al. · stanford

Recent advancements in autonomous driving have relied on data-driven approaches, which are widely adopted but face challenges including dataset bias, overfitting, and uninterpretability. Drawing inspiration from the knowledge-driven nature of human driving, we explore the question of how to instill similar capabilities into autonomous driving systems and summarize a paradigm that integrates an interactive environment, a driver agent, as well as a memory component to address this question. Leveraging large language models (LLMs) with emergent abilities, we propose the DiLu framework, which combines a Reasoning and a Reflection module to enable the system to perform decision-making based on common-sense knowledge and evolve continuously. Extensive experiments prove DiLu's capability to accumulate experience and demonstrate a significant advantage in generalization ability over reinforcement learning-based methods. Moreover, DiLu is able to directly acquire experiences from real-world datasets which highlights its potential to be deployed on practical autonomous driving systems. To the best of our knowledge, we are the first to leverage knowledge-driven capability in decision-making for autonomous vehicles. Through the proposed DiLu framework, LLM is strengthened to apply knowledge and to reason causally in the autonomous driving domain. Project page: https://pjlab-adg.github.io/DiLu/

Ben Fei, Weidong Yang, Wenming Chen et al. · stanford

Point cloud completion is a generation and estimation issue derived from the partial point clouds, which plays a vital role in the applications in 3D computer vision. The progress of deep learning (DL) has impressively improved the capability and robustness of point cloud completion. However, the quality of completed point clouds is still needed to be further enhanced to meet the practical utilization. Therefore, this work aims to conduct a comprehensive survey on various methods, including point-based, convolution-based, graph-based, and generative model-based approaches, etc. And this survey summarizes the comparisons among these methods to provoke further research insights. Besides, this review sums up the commonly used datasets and illustrates the applications of point cloud completion. Eventually, we also discussed possible research trends in this promptly expanding field.

Tao Ma, Xuemeng Yang, Hongbin Zhou et al. · stanford

Existing offboard 3D detectors always follow a modular pipeline design to take advantage of unlimited sequential point clouds. We have found that the full potential of offboard 3D detectors is not explored mainly due to two reasons: (1) the onboard multi-object tracker cannot generate sufficient complete object trajectories, and (2) the motion state of objects poses an inevitable challenge for the object-centric refining stage in leveraging the long-term temporal context representation. To tackle these problems, we propose a novel paradigm of offboard 3D object detection, named DetZero. Concretely, an offline tracker coupled with a multi-frame detector is proposed to focus on the completeness of generated object tracks. An attention-mechanism refining module is proposed to strengthen contextual information interaction across long-term sequential point clouds for object refining with decomposed regression methods. Extensive experiments on Waymo Open Dataset show our DetZero outperforms all state-of-the-art onboard and offboard 3D detection methods. Notably, DetZero ranks 1st place on Waymo 3D object detection leaderboard with 85.15 mAPH (L2) detection performance. Further experiments validate the application of taking the place of human labels with such high-quality results. Our empirical study leads to rethinking conventions and interesting findings that can guide future research on offboard 3D object detection.

Zheng Tong, Tao Ma, Ju Huyan et al. · stanford

Pavement damage segmentation has benefited enormously from deep learning. % and large-scale datasets. However, few current public datasets limit the potential exploration of deep learning in the application of pavement damage segmentation. To address this problem, this study has proposed Pavementscapes, a large-scale dataset to develop and evaluate methods for pavement damage segmentation. Pavementscapes is comprised of 4,000 images with a resolution of $1024 \times 2048$, which have been recorded in the real-world pavement inspection projects with 15 different pavements. A total of 8,680 damage instances are manually labeled with six damage classes at the pixel level. The statistical study gives a thorough investigation and analysis of the proposed dataset. The numeral experiments propose the top-performing deep neural networks capable of segmenting pavement damages, which provides the baselines of the open challenge for pavement inspection. The experiment results also indicate the existing problems for damage segmentation using deep learning, and this study provides potential solutions.

Licheng Wen, Xuemeng Yang, Daocheng Fu et al.

The pursuit of autonomous driving technology hinges on the sophisticated integration of perception, decision-making, and control systems. Traditional approaches, both data-driven and rule-based, have been hindered by their inability to grasp the nuance of complex driving environments and the intentions of other road users. This has been a significant bottleneck, particularly in the development of common sense reasoning and nuanced scene understanding necessary for safe and reliable autonomous driving. The advent of Visual Language Models (VLM) represents a novel frontier in realizing fully autonomous vehicle driving. This report provides an exhaustive evaluation of the latest state-of-the-art VLM, GPT-4V(ision), and its application in autonomous driving scenarios. We explore the model's abilities to understand and reason about driving scenes, make decisions, and ultimately act in the capacity of a driver. Our comprehensive tests span from basic scene recognition to complex causal reasoning and real-time decision-making under varying conditions. Our findings reveal that GPT-4V demonstrates superior performance in scene understanding and causal reasoning compared to existing autonomous systems. It showcases the potential to handle out-of-distribution scenarios, recognize intentions, and make informed decisions in real driving contexts. However, challenges remain, particularly in direction discernment, traffic light recognition, vision grounding, and spatial reasoning tasks. These limitations underscore the need for further research and development. Project is now available on GitHub for interested parties to access and utilize: \url{https://github.com/PJLab-ADG/GPT4V-AD-Exploration}

Tao Ma, Jin Zhu, Hengrui Cai et al.

In real-world applications of reinforcement learning, it is often challenging to obtain a state representation that is parsimonious and satisfies the Markov property without prior knowledge. Consequently, it is common practice to construct a state larger than necessary, e.g., by concatenating measurements over contiguous time points. However, needlessly increasing the dimension of the state may slow learning and obfuscate the learned policy. We introduce the notion of a minimal sufficient state in a Markov decision process (MDP) as the subvector of the original state under which the process remains an MDP and shares the same reward function as the original process. We propose a novel SEquEntial Knockoffs (SEEK) algorithm that estimates the minimal sufficient state in a system with high-dimensional complex nonlinear dynamics. In large samples, the proposed method achieves selection consistency. As the method is agnostic to the reinforcement learning algorithm being applied, it benefits downstream tasks such as policy learning. Empirical experiments verify theoretical results and show the proposed approach outperforms several competing methods regarding variable selection accuracy and regret.

Xin Li, Yeqi Bai, Pinlong Cai et al.

This paper explores the emerging knowledge-driven autonomous driving technologies. Our investigation highlights the limitations of current autonomous driving systems, in particular their sensitivity to data bias, difficulty in handling long-tail scenarios, and lack of interpretability. Conversely, knowledge-driven methods with the abilities of cognition, generalization and life-long learning emerge as a promising way to overcome these challenges. This paper delves into the essence of knowledge-driven autonomous driving and examines its core components: dataset \& benchmark, environment, and driver agent. By leveraging large language models, world models, neural rendering, and other advanced artificial intelligence techniques, these components collectively contribute to a more holistic, adaptive, and intelligent autonomous driving system. The paper systematically organizes and reviews previous research efforts in this area, and provides insights and guidance for future research and practical applications of autonomous driving. We will continually share the latest updates on cutting-edge developments in knowledge-driven autonomous driving along with the relevant valuable open-source resources at: \url{https://github.com/PJLab-ADG/awesome-knowledge-driven-AD}.

Jialiang Huang, Teng Ma, Zheng Liu et al.

Serverless computing is renowned for its computation elasticity, yet its full potential is often constrained by the requirement for functions to operate within local and dedicated background environments, resulting in limited memory elasticity. To address this limitation, this paper introduces TrEnv-X, a co-designed integration of the serverless platform with the operating system and CXL/RDMA-based remote memory pools. TrEnv-X's core innovations are repurposable sandboxes, which can be shared across different functions to decrease the associated creation overhead, and OS-level memory templates, which enable rapid state restoration from CXL/RDMA-based remote memory pools. To further demonstrate TrEnv-X's versatility, we generalize its design from traditional containers for microVM-based agent workloads and introduce new optimizations, including browser sharing and a page cache bypassing mechanism. Our evaluation shows that TrEnv-X achieves up to 7x reduction in P99 latency and 48% memory savings for container-based functions. When applied to LLM agents, it reduces the P99 latency by up to 58% and memory usage by 61% compared to state-of-the-art systems like E2B.

Yusheng Zheng, Wenan Mao, Shuyi Cheng et al.

Performance diagnosis in production-scale AI training is challenging because subtle OS-level issues can trigger cascading GPU delays and network slowdowns, degrading training efficiency across thousands of GPUs. Existing profiling tools are limited to single system layers, incur prohibitive overhead (10--30%), or lack continuous deployment capabilities, resulting in manual analyses spanning days. We argue that continuous, cross-layer observability enabled by OS-level instrumentation and layered differential diagnosis is necessary to address this gap. We introduce SysOM-AI, a production observability system that continuously integrates CPU stack profiling, GPU kernel tracing, and NCCL event instrumentation via adaptive hybrid stack unwinding and eBPF-based tracing, incurring less than 0.4% overhead. Deployed at Alibaba across over 80,000 GPUs for more than one year, SysOM-AI helped diagnose 94 confirmed production issues, reducing median diagnosis time from days to approximately 10 minutes.

Tao Ma, Chao Zhang, Min Lu et al.

Renal pathology, as the gold standard of kidney disease diagnosis, requires doctors to analyze a series of tissue slices stained by H&E staining and special staining like Masson, PASM, and PAS, respectively. These special staining methods are costly, time-consuming, and hard to standardize for wide use especially in primary hospitals. Advances of supervised learning methods have enabled the virtually conversion of H&E images into special staining images, but achieving pixel-to-pixel alignment for training remains challenging. In contrast, unsupervised learning methods regarding different stains as different style transfer domains can utilize unpaired data, but they ignore the spatial inter-domain correlations and thus decrease the trustworthiness of structural details for diagnosis. In this paper, we propose a novel virtual staining framework AGMDT to translate images into other domains by avoiding pixel-level alignment and meanwhile utilizing the correlations among adjacent tissue slices. We first build a high-quality multi-domain renal histological dataset where each specimen case comprises a series of slices stained in various ways. Based on it, the proposed framework AGMDT discovers patch-level aligned pairs across the serial slices of multi-domains through glomerulus detection and bipartite graph matching, and utilizes such correlations to supervise the end-to-end model for multi-domain staining transformation. Experimental results show that the proposed AGMDT achieves a good balance between the precise pixel-level alignment and unpaired domain transfer by exploiting correlations across multi-domain serial pathological slices, and outperforms the state-of-the-art methods in both quantitative measure and morphological details.

Tao Ma, Xuzhi Yang, Zoltan Szabo

Reinforcement learning (RL) -- finding the optimal behaviour (also referred to as policy) maximizing the collected long-term cumulative reward -- is among the most influential approaches in machine learning with a large number of successful applications. In several decision problems, however, one faces the possibility of policy switching -- changing from the current policy to a new one -- which incurs a non-negligible cost, and in the decision one is limited to using historical data without the availability for further online interaction. Despite the inevitable importance of this offline learning scenario, to our best knowledge, very little effort has been made to tackle the key problem of balancing between the gain and the cost of switching in a flexible and principled way. Leveraging ideas from the area of optimal transport, we initialize the systematic study of policy switching in offline RL. We establish fundamental properties and design a Net Actor-Critic algorithm for the proposed novel switching formulation. Numerical experiments demonstrate the efficiency of our approach on multiple robot control benchmarks of the Gymnasium and traffic light control from SUMO-RL.

Davide Ghilardi, Federico Belotti, Marco Molinari et al.

SAEs have recently been employed as a promising unsupervised approach for understanding the representations of layers of Large Language Models (LLMs). However, with the growth in model size and complexity, training SAEs is computationally intensive, as typically one SAE is trained for each model layer. To address such limitation, we propose \textit{Group-SAE}, a novel strategy to train SAEs. Our method considers the similarity of the residual stream representations between contiguous layers to group similar layers and train a single SAE per group. To balance the trade-off between efficiency and performance, we further introduce \textit{AMAD} (Average Maximum Angular Distance), an empirical metric that guides the selection of an optimal number of groups based on representational similarity across layers. Experiments on models from the Pythia family show that our approach significantly accelerates training with minimal impact on reconstruction quality and comparable downstream task performance and interpretability over baseline SAEs trained layer by layer. This method provides an efficient and scalable strategy for training SAEs in modern LLMs.

Philipp Schoenegger, Francesco Salvi, Jiacheng Liu et al. · oxford

We directly compare the persuasion capabilities of a frontier large language model (LLM; Claude Sonnet 3.5) against incentivized human persuaders in an interactive, real-time conversational quiz setting. In this preregistered, large-scale incentivized experiment, participants (quiz takers) completed an online quiz where persuaders (either humans or LLMs) attempted to persuade quiz takers toward correct or incorrect answers. We find that LLM persuaders achieved significantly higher compliance with their directional persuasion attempts than incentivized human persuaders, demonstrating superior persuasive capabilities in both truthful (toward correct answers) and deceptive (toward incorrect answers) contexts. We also find that LLM persuaders significantly increased quiz takers' accuracy, leading to higher earnings, when steering quiz takers toward correct answers, and significantly decreased their accuracy, leading to lower earnings, when steering them toward incorrect answers. Overall, our findings suggest that AI's persuasion capabilities already exceed those of humans that have real-money bonuses tied to performance. Our findings of increasingly capable AI persuaders thus underscore the urgency of emerging alignment and governance frameworks.

Xuesong Chen, Linjiang Huang, Tao Ma et al.

The integration of Vision-Language Models (VLMs) into autonomous driving systems has shown promise in addressing key challenges such as learning complexity, interpretability, and common-sense reasoning. However, existing approaches often struggle with efficient integration and realtime decision-making due to computational demands. In this paper, we introduce SOLVE, an innovative framework that synergizes VLMs with end-to-end (E2E) models to enhance autonomous vehicle planning. Our approach emphasizes knowledge sharing at the feature level through a shared visual encoder, enabling comprehensive interaction between VLM and E2E components. We propose a Trajectory Chain-of-Thought (T-CoT) paradigm, which progressively refines trajectory predictions, reducing uncertainty and improving accuracy. By employing a temporal decoupling strategy, SOLVE achieves efficient cooperation by aligning high-quality VLM outputs with E2E real-time performance. Evaluated on the nuScenes dataset, our method demonstrates significant improvements in trajectory prediction accuracy, paving the way for more robust and reliable autonomous driving systems.

Xuesong Chen, Shaoshuai Shi, Tao Ma et al.

The perception system for autonomous driving generally requires to handle multiple diverse sub-tasks. However, current algorithms typically tackle individual sub-tasks separately, which leads to low efficiency when aiming at obtaining full-perception results. Some multi-task learning methods try to unify multiple tasks with one model, but do not solve the conflicts in multi-task learning. In this paper, we introduce M3Net, a novel multimodal and multi-task network that simultaneously tackles detection, segmentation, and 3D occupancy prediction for autonomous driving and achieves superior performance than single task model. M3Net takes multimodal data as input and multiple tasks via query-token interactions. To enhance the integration of multi-modal features for multi-task learning, we first propose the Modality-Adaptive Feature Integration (MAFI) module, which enables single-modality features to predict channel-wise attention weights for their high-performing tasks, respectively. Based on integrated features, we then develop task-specific query initialization strategies to accommodate the needs of detection/segmentation and 3D occupancy prediction. Leveraging the properly initialized queries, a shared decoder transforms queries and BEV features layer-wise, facilitating multi-task learning. Furthermore, we propose a Task-oriented Channel Scaling (TCS) module in the decoder to mitigate conflicts between optimizing for different tasks. Additionally, our proposed multi-task querying and TCS module support both Transformer-based decoder and Mamba-based decoder, demonstrating its flexibility to different architectures. M3Net achieves state-of-the-art multi-task learning performance on the nuScenes benchmarks.

Tao Ma, Hongbin Zhou, Qiusheng Huang et al.

Offboard perception aims to automatically generate high-quality 3D labels for autonomous driving (AD) scenes. Existing offboard methods focus on 3D object detection with closed-set taxonomy and fail to match human-level recognition capability on the rapidly evolving perception tasks. Due to heavy reliance on human labels and the prevalence of data imbalance and sparsity, a unified framework for offboard auto-labeling various elements in AD scenes that meets the distinct needs of perception tasks is not being fully explored. In this paper, we propose a novel multi-modal Zero-shot Offboard Panoptic Perception (ZOPP) framework for autonomous driving scenes. ZOPP integrates the powerful zero-shot recognition capabilities of vision foundation models and 3D representations derived from point clouds. To the best of our knowledge, ZOPP represents a pioneering effort in the domain of multi-modal panoptic perception and auto labeling for autonomous driving scenes. We conduct comprehensive empirical studies and evaluations on Waymo open dataset to validate the proposed ZOPP on various perception tasks. To further explore the usability and extensibility of our proposed ZOPP, we also conduct experiments in downstream applications. The results further demonstrate the great potential of our ZOPP for real-world scenarios.

Siqi Wang, Zehui Zhu, Tao Ma et al.

Digital Image Correlation (DIC) is an optical technique that measures displacement and strain by tracking pattern movement in a sequence of captured images during testing. DIC has gained recognition in asphalt pavement engineering since the early 2000s. However, users often perceive the DIC technique as an out-of-box tool and lack a thorough understanding of its operational and measurement principles. This article presents a state-of-art review of DIC as a crucial tool for laboratory testing of asphalt concrete (AC), primarily focusing on the widely utilized 2D-DIC and 3D-DIC techniques. To address frequently asked questions from users, the review thoroughly examines the optimal methods for preparing speckle patterns, configuring single-camera or dual-camera imaging systems, conducting DIC analyses, and exploring various applications. Furthermore, emerging DIC methodologies such as Digital Volume Correlation and deep-learning-based DIC are introduced, highlighting their potential for future applications in pavement engineering. The article also provides a comprehensive and reliable flowchart for implementing DIC in AC characterization. Finally, critical directions for future research are presented.

Xueying Du, Geng Zheng, Kaixin Wang et al.

Although LLMs have shown promising potential in vulnerability detection, this study reveals their limitations in distinguishing between vulnerable and similar-but-benign patched code (only 0.06 - 0.14 accuracy). It shows that LLMs struggle to capture the root causes of vulnerabilities during vulnerability detection. To address this challenge, we propose enhancing LLMs with multi-dimensional vulnerability knowledge distilled from historical vulnerabilities and fixes. We design a novel knowledge-level Retrieval-Augmented Generation framework Vul-RAG, which improves LLMs with an accuracy increase of 16% - 24% in identifying vulnerable and patched code. Additionally, vulnerability knowledge generated by Vul-RAG can further (1) serve as high-quality explanations to improve manual detection accuracy (from 60% to 77%), and (2) detect 10 previously-unknown bugs in the recent Linux kernel release with 6 assigned CVEs.

Tao Ma, Yikang Li

Generating images with conditional descriptions gains increasing interests in recent years. However, existing conditional inputs are suffering from either unstructured forms (captions) or limited information and expensive labeling (scene graphs). For a targeted scene, the core items, objects, are usually definite while their interactions are flexible and hard to clearly define. Thus, we introduce a more rational setting, generating a realistic image from the objects and captions. Under this setting, objects explicitly define the critical roles in the targeted images and captions implicitly describe their rich attributes and connections. Correspondingly, a MOC-GAN is proposed to mix the inputs of two modalities to generate realistic images. It firstly infers the implicit relations between object pairs from the captions to build a hidden-state scene graph. So a multi-layer representation containing objects, relations and captions is constructed, where the scene graph provides the structures of the scene and the caption provides the image-level guidance. Then a cascaded attentive generative network is designed to coarse-to-fine generate phrase patch by paying attention to the most relevant words in the caption. In addition, a phrase-wise DAMSM is proposed to better supervise the fine-grained phrase-patch consistency. On COCO dataset, our method outperforms the state-of-the-art methods on both Inception Score and FID while maintaining high visual quality. Extensive experiments demonstrate the unique features of our proposed method.

Tao Ma, Zhizheng Liu, Yikang Li

Sensor configuration, including the sensor selections and their installation locations, serves a crucial role in autonomous driving. A well-designed sensor configuration significantly improves the performance upper bound of the perception system. However, as leveraging multiple sensors is becoming the mainstream setting, existing methods mainly focusing on single-sensor configuration problems are hardly utilized in practice. To tackle these issues, we propose a novel method based on conditional entropy in Bayesian theory to evaluate the sensor configurations containing both cameras and LiDARs. Correspondingly, an evaluation metric, perception entropy, is introduced to measure the difference between two configurations, which considers both the perception algorithm performance and the selections of the sensors. To the best of our knowledge, this is the first method to tackle the multi-sensor configuration problem for autonomous vehicles. The simulation results, extensive comparisons, and analysis all demonstrate the superior performance of our proposed approach.

Tao Ma, Zhizheng Liu, Guohang Yan et al.

For autonomous vehicles, an accurate calibration for LiDAR and camera is a prerequisite for multi-sensor perception systems. However, existing calibration techniques require either a complicated setting with various calibration targets, or an initial calibration provided beforehand, which greatly impedes their applicability in large-scale autonomous vehicle deployment. To tackle these issues, we propose a novel method to calibrate the extrinsic parameter for LiDAR and camera in road scenes. Our method introduces line features from static straight-line-shaped objects such as road lanes and poles in both image and point cloud and formulates the initial calibration of extrinsic parameters as a perspective-3-lines (P3L) problem. Subsequently, a cost function defined under the semantic constraints of the line features is designed to perform refinement on the solved coarse calibration. The whole procedure is fully automatic and user-friendly without the need to adjust environment settings or provide an initial calibration. We conduct extensive experiments on KITTI and our in-house dataset, quantitative and qualitative results demonstrate the robustness and accuracy of our method.

Yeqi Bai, Tao Ma, Lipo Wang et al.

While deep learning technologies are now capable of generating realistic images confusing humans, the research efforts are turning to the synthesis of images for more concrete and application-specific purposes. Facial image generation based on vocal characteristics from speech is one of such important yet challenging tasks. It is the key enabler to influential use cases of image generation, especially for business in public security and entertainment. Existing solutions to the problem of speech2face renders limited image quality and fails to preserve facial similarity due to the lack of quality dataset for training and appropriate integration of vocal features. In this paper, we investigate these key technical challenges and propose Speech Fusion to Face, or SF2F in short, attempting to address the issue of facial image quality and the poor connection between vocal feature domain and modern image generation models. By adopting new strategies on data model and training, we demonstrate dramatic performance boost over state-of-the-art solution, by doubling the recall of individual identity, and lifting the quality score from 15 to 19 based on the mutual information score with VGGFace classifier.

Kyu J. Han, Jing Pan, Venkata Krishna Naveen Tadala et al.

This paper proposes multistream CNN, a novel neural network architecture for robust acoustic modeling in speech recognition tasks. The proposed architecture processes input speech with diverse temporal resolutions by applying different dilation rates to convolutional neural networks across multiple streams to achieve the robustness. The dilation rates are selected from the multiples of a sub-sampling rate of 3 frames. Each stream stacks TDNN-F layers (a variant of 1D CNN), and output embedding vectors from the streams are concatenated then projected to the final layer. We validate the effectiveness of the proposed multistream CNN architecture by showing consistent improvements against Kaldi's best TDNN-F model across various data sets. Multistream CNN improves the WER of the test-other set in the LibriSpeech corpus by 12% (relative). On custom data from ASAPP's production ASR system for a contact center, it records a relative WER improvement of 11% for customer channel audio to prove its robustness to data in the wild. In terms of real-time factor, multistream CNN outperforms the baseline TDNN-F by 15%, which also suggests its practicality on production systems. When combined with self-attentive SRU LM rescoring, multistream CNN contributes for ASAPP to achieve the best WER of 1.75% on test-clean in LibriSpeech.

Jing Pan, Joshua Shapiro, Jeremy Wohlwend et al.

In this paper we present state-of-the-art (SOTA) performance on the LibriSpeech corpus with two novel neural network architectures, a multistream CNN for acoustic modeling and a self-attentive simple recurrent unit (SRU) for language modeling. In the hybrid ASR framework, the multistream CNN acoustic model processes an input of speech frames in multiple parallel pipelines where each stream has a unique dilation rate for diversity. Trained with the SpecAugment data augmentation method, it achieves relative word error rate (WER) improvements of 4% on test-clean and 14% on test-other. We further improve the performance via N-best rescoring using a 24-layer self-attentive SRU language model, achieving WERs of 1.75% on test-clean and 4.46% on test-other.

Kyu J. Han, Ramon Prieto, Kaixing Wu et al.

Self-attention has been a huge success for many downstream tasks in NLP, which led to exploration of applying self-attention to speech problems as well. The efficacy of self-attention in speech applications, however, seems not fully blown yet since it is challenging to handle highly correlated speech frames in the context of self-attention. In this paper we propose a new neural network model architecture, namely multi-stream self-attention, to address the issue thus make the self-attention mechanism more effective for speech recognition. The proposed model architecture consists of parallel streams of self-attention encoders, and each stream has layers of 1D convolutions with dilated kernels whose dilation rates are unique given stream, followed by a self-attention layer. The self-attention mechanism in each stream pays attention to only one resolution of input speech frames and the attentive computation can be more efficient. In a later stage, outputs from all the streams are concatenated then linearly projected to the final embedding. By stacking the proposed multi-stream self-attention encoder blocks and rescoring the resultant lattices with neural network language models, we achieve the word error rate of 2.2% on the test-clean dataset of the LibriSpeech corpus, the best number reported thus far on the dataset.

Yikang Li, Tao Ma, Yeqi Bai et al.

Despite some exciting progress on high-quality image generation from structured(scene graphs) or free-form(sentences) descriptions, most of them only guarantee the image-level semantical consistency, i.e. the generated image matching the semantic meaning of the description. They still lack the investigations on synthesizing the images in a more controllable way, like finely manipulating the visual appearance of every object. Therefore, to generate the images with preferred objects and rich interactions, we propose a semi-parametric method, PasteGAN, for generating the image from the scene graph and the image crops, where spatial arrangements of the objects and their pair-wise relationships are defined by the scene graph and the object appearances are determined by the given object crops. To enhance the interactions of the objects in the output, we design a Crop Refining Network and an Object-Image Fuser to embed the objects as well as their relationships into one map. Multiple losses work collaboratively to guarantee the generated images highly respecting the crops and complying with the scene graphs while maintaining excellent image quality. A crop selector is also proposed to pick the most-compatible crops from our external object tank by encoding the interactions around the objects in the scene graph if the crops are not provided. Evaluated on Visual Genome and COCO-Stuff dataset, our proposed method significantly outperforms the SOTA methods on Inception Score, Diversity Score and Fréchet Inception Distance. Extensive experiments also demonstrate our method's ability to generate complex and diverse images with given objects.

Tao Ma

Financial forecasting is challenging and attractive in machine learning. There are many classic solutions, as well as many deep learning based methods, proposed to deal with it yielding encouraging performance. Stock time series forecasting is the most representative problem in financial forecasting. Due to the strong connections among stocks, the information valuable for forecasting is not only included in individual stocks, but also included in the stocks related to them. However, most previous works focus on one single stock, which easily ignore the valuable information in others. To leverage more information, in this paper, we propose a jointly forecasting approach to process multiple time series of related stocks simultaneously, using multi-task learning framework. Compared to the previous works, we use multiple networks to forecast multiple related stocks, using the shared and private information of them simultaneously through multi-task learning. Moreover, we propose an attention method learning an optimized weighted combination of shared and private information based on the idea of Capital Asset Pricing Model (CAPM) to help forecast. Experimental results on various data show improved forecasting performance over baseline methods.