Yu Gu

Yu Gu, Jianwei Yang, Naoto Usuyama et al. · microsoft-research

Rapid progress has been made in instruction-learning for image editing with natural-language instruction, as exemplified by InstructPix2Pix. In biomedicine, such methods can be applied to counterfactual image generation, which helps differentiate causal structure from spurious correlation and facilitate robust image interpretation for disease progression modeling. However, generic image-editing models are ill-suited for the biomedical domain, and counterfactual biomedical image generation is largely underexplored. In this paper, we present BiomedJourney, a novel method for counterfactual biomedical image generation by instruction-learning from multimodal patient journeys. Given a patient with two biomedical images taken at different time points, we use GPT-4 to process the corresponding imaging reports and generate a natural language description of disease progression. The resulting triples (prior image, progression description, new image) are then used to train a latent diffusion model for counterfactual biomedical image generation. Given the relative scarcity of image time series data, we introduce a two-stage curriculum that first pretrains the denoising network using the much more abundant single image-report pairs (with dummy prior image), and then continues training using the counterfactual triples. Experiments using the standard MIMIC-CXR dataset demonstrate the promise of our method. In a comprehensive battery of tests on counterfactual medical image generation, BiomedJourney substantially outperforms prior state-of-the-art methods in instruction image editing and medical image generation such as InstructPix2Pix and RoentGen. To facilitate future study in counterfactual medical generation, we plan to release our instruction-learning code and pretrained models.

Zhenghao Liu, Sen Mei, Chenyan Xiong et al.

This paper proposes Text mAtching based SequenTial rEcommendation model (TASTE), which maps items and users in an embedding space and recommends items by matching their text representations. TASTE verbalizes items and user-item interactions using identifiers and attributes of items. To better characterize user behaviors, TASTE additionally proposes an attention sparsity method, which enables TASTE to model longer user-item interactions by reducing the self-attention computations during encoding. Our experiments show that TASTE outperforms the state-of-the-art methods on widely used sequential recommendation datasets. TASTE alleviates the cold start problem by representing long-tail items using full-text modeling and bringing the benefits of pretrained language models to recommendation systems. Our further analyses illustrate that TASTE significantly improves the recommendation accuracy by reducing the popularity bias of previous item id based recommendation models and returning more appropriate and text-relevant items to satisfy users. All codes are available at https://github.com/OpenMatch/TASTE.

Yumeng Song, Yu Gu, Tianyi Li et al.

Hypergraphs can model higher-order relationships among data objects that are found in applications such as social networks and bioinformatics. However, recent studies on hypergraph learning that extend graph convolutional networks to hypergraphs cannot learn effectively from features of unlabeled data. To such learning, we propose a contrastive hypergraph neural network, CHGNN, that exploits self-supervised contrastive learning techniques to learn from labeled and unlabeled data. First, CHGNN includes an adaptive hypergraph view generator that adopts an auto-augmentation strategy and learns a perturbed probability distribution of minimal sufficient views. Second, CHGNN encompasses an improved hypergraph encoder that considers hyperedge homogeneity to fuse information effectively. Third, CHGNN is equipped with a joint loss function that combines a similarity loss for the view generator, a node classification loss, and a hyperedge homogeneity loss to inject supervision signals. It also includes basic and cross-validation contrastive losses, associated with an enhanced contrastive loss training process. Experimental results on nine real datasets offer insight into the effectiveness of CHGNN, showing that it outperforms 13 competitors in terms of classification accuracy consistently.

Marcela Mera-Trujillo, Shivang Patel, Yu Gu et al.

Keypoint detection and matching is a fundamental task in many computer vision problems, from shape reconstruction, to structure from motion, to AR/VR applications and robotics. It is a well-studied problem with remarkable successes such as SIFT, and more recent deep learning approaches. While great robustness is exhibited by these techniques with respect to noise, illumination variation, and rigid motion transformations, less attention has been placed on image distortion sensitivity. In this work, we focus on the case when this is caused by the geometry of the cameras used for image acquisition, and consider the keypoint detection and matching problem between the hybrid scenario of a fisheye and a projective image. We build on a state-of-the-art approach and derive a self-supervised procedure that enables training an interest point detector and descriptor network. We also collected two new datasets for additional training and testing in this unexplored scenario, and we demonstrate that current approaches are suboptimal because they are designed to work in traditional projective conditions, while the proposed approach turns out to be the most effective.

Weipan Xu, Yu Gu, Yifan Chen et al.

Housing quality is an essential proxy for regional wealth, security and health. Understanding the distribution of housing quality is crucial for unveiling rural development status and providing political proposals. However,present rural house quality data highly depends on a top-down, time-consuming survey at the national or provincial level but fails to unpack the housing quality at the village level. To fill the gap between accurately depicting rural housing quality conditions and deficient data,we collect massive rural images and invite users to assess their housing quality at scale. Furthermore, a deep learning framework is proposed to automatically and efficiently predict housing quality based on crowd-sourcing rural images.

Manjie Xu, Chenxing Li, Xinyi Tu et al.

Generating semantically and temporally aligned audio content in accordance with video input has become a focal point for researchers, particularly following the remarkable breakthrough in text-to-video generation. In this work, we aim to offer insights into the video-to-audio generation paradigm, focusing on three crucial aspects: vision encoders, auxiliary embeddings, and data augmentation techniques. Beginning with a foundational model built on a simple yet surprisingly effective intuition, we explore various vision encoders and auxiliary embeddings through ablation studies. Employing a comprehensive evaluation pipeline that emphasizes generation quality and video-audio synchronization alignment, we demonstrate that our model exhibits state-of-the-art video-to-audio generation capabilities. Furthermore, we provide critical insights into the impact of different data augmentation methods on enhancing the generation framework's overall capacity. We showcase possibilities to advance the challenge of generating synchronized audio from semantic and temporal perspectives. We hope these insights will serve as a stepping stone toward developing more realistic and accurate audio-visual generation models.

Heng Rao, Yu Gu, Jason Zipeng Zhang et al.

Biological oscillations are periodic changes in various signaling processes crucial for the proper functioning of living organisms. These oscillations are modeled by ordinary differential equations, with coefficient variations leading to diverse periodic behaviors, typically measured by oscillatory frequencies. This paper explores sampling techniques for neural networks to model the relationship between system coefficients and oscillatory frequency. However, the scarcity of oscillations in the vast coefficient space results in many samples exhibiting non-periodic behaviors, and small coefficient changes near oscillation boundaries can significantly alter oscillatory properties. This leads to non-oscillatory bias and boundary sensitivity, making accurate predictions difficult. While existing importance and uncertainty sampling approaches partially mitigate these challenges, they either fail to resolve the sensitivity problem or result in redundant sampling. To address these limitations, we propose the Hierarchical Gradient-based Genetic Sampling (HGGS) framework, which improves the accuracy of neural network predictions for biological oscillations. The first layer, Gradient-based Filtering, extracts sensitive oscillation boundaries and removes redundant non-oscillatory samples, creating a balanced coarse dataset. The second layer, Multigrid Genetic Sampling, utilizes residual information to refine these boundaries and explore new high-residual regions, increasing data diversity for model training. Experimental results demonstrate that HGGS outperforms seven comparative sampling methods across four biological systems, highlighting its effectiveness in enhancing sampling and prediction accuracy.

Ram Zaveri, Shivang Patel, Yu Gu et al.

Efficient visual trackers overfit to their training distributions and lack generalization abilities, resulting in them performing well on their respective in-distribution (ID) test sets and not as well on out-of-distribution (OOD) sequences, imposing limitations to their deployment in-the-wild under constrained resources. We introduce SiamABC, a highly efficient Siamese tracker that significantly improves tracking performance, even on OOD sequences. SiamABC takes advantage of new architectural designs in the way it bridges the dynamic variability of the target, and of new losses for training. Also, it directly addresses OOD tracking generalization by including a fast backward-free dynamic test-time adaptation method that continuously adapts the model according to the dynamic visual changes of the target. Our extensive experiments suggest that SiamABC shows remarkable performance gains in OOD sets while maintaining accurate performance on the ID benchmarks. SiamABC outperforms MixFormerV2-S by 7.6\% on the OOD AVisT benchmark while being 3x faster (100 FPS) on a CPU. Our code and models are available at https://wvuvl.github.io/SiamABC/.

Xiang Zhang, Yan Lu, Huan Yan et al.

Because of the ambiguous and subjective property of the facial expression recognition (FER) task, the label noise is widely existing in the FER dataset. For this problem, in the training phase, current FER methods often directly predict whether the label of the input image is noised or not, aiming to reduce the contribution of the noised data in training. However, we argue that this kind of method suffers from the low reliability of such noise data decision operation. It makes that some mistakenly abounded clean data are not utilized sufficiently and some mistakenly kept noised data disturbing the model learning process. In this paper, we propose a more reliable noise-label suppression method called ReSup (Reliable label noise Suppression for FER). First, instead of directly predicting noised or not, ReSup makes the noise data decision by modeling the distribution of noise and clean labels simultaneously according to the disagreement between the prediction and the target. Specifically, to achieve optimal distribution modeling, ReSup models the similarity distribution of all samples. To further enhance the reliability of our noise decision results, ReSup uses two networks to jointly achieve noise suppression. Specifically, ReSup utilize the property that two networks are less likely to make the same mistakes, making two networks swap decisions and tending to trust decisions with high agreement. Extensive experiments on three popular benchmarks show that the proposed method significantly outperforms state-of-the-art noisy label FER methods by 3.01% on FERPlus becnmarks. Code: https://github.com/purpleleaves007/FERDenoise

Camndon Reed, Christopher Tatsch, Jason N. Gross et al.

Natural environments pose significant challenges for autonomous robot navigation, particularly due to their unstructured and ever-changing nature. Hiking trails, with their dynamic conditions influenced by weather, vegetation, and human traffic, represent one such challenge. This work introduces a novel approach to autonomous hiking trail navigation that balances trail adherence with the flexibility to adapt to off-trail routes when necessary. The solution is a Traversability Analysis module that integrates semantic data from camera images with geometric information from LiDAR to create a comprehensive understanding of the surrounding terrain. A planner uses this traversability map to navigate safely, adhering to trails while allowing off-trail movement when necessary to avoid on-trail hazards or for safe off-trail shortcuts. The method is evaluated through simulation to determine the balance between semantic and geometric information in traversability estimation. These simulations tested various weights to assess their impact on navigation performance across different trail scenarios. Weights were then validated through field tests at the West Virginia University Core Arboretum, demonstrating the method's effectiveness in a real-world environment.

Jianwei Cui, Yu Gu, Chao Weng et al.

This paper presents an advanced end-to-end singing voice synthesis (SVS) system based on the source-filter mechanism that directly translates lyrical and melodic cues into expressive and high-fidelity human-like singing. Similarly to VISinger 2, the proposed system also utilizes training paradigms evolved from VITS and incorporates elements like the fundamental pitch (F0) predictor and waveform generation decoder. To address the issue that the coupling of mel-spectrogram features with F0 information may introduce errors during F0 prediction, we consider two strategies. Firstly, we leverage mel-cepstrum (mcep) features to decouple the intertwined mel-spectrogram and F0 characteristics. Secondly, inspired by the neural source-filter models, we introduce source excitation signals as the representation of F0 in the SVS system, aiming to capture pitch nuances more accurately. Meanwhile, differentiable mcep and F0 losses are employed as the waveform decoder supervision to fortify the prediction accuracy of speech envelope and pitch in the generated speech. Experiments on the Opencpop dataset demonstrate efficacy of the proposed model in synthesis quality and intonation accuracy.

Jiexuan Zhang, Yiheng Du, Qian Wang et al.

Despite their generative power, diffusion models struggle to maintain style consistency across images conditioned on the same style prompt, hindering their practical deployment in creative workflows. While several training-free methods attempt to solve this, they are constrained to the U-Net architecture, which not only leads to low-quality results and artifacts like object repetition but also renders them incompatible with superior Diffusion Transformer (DiT). To address these issues, we introduce AlignedGen, a novel training-free framework that enhances style consistency across images generated by DiT models. Our work first reveals a critical insight: naive attention sharing fails in DiT due to conflicting positional signals from improper position embeddings. We introduce Shifted Position Embedding (ShiftPE), an effective solution that resolves this conflict by allocating a non-overlapping set of positional indices to each image. Building on this foundation, we develop Advanced Attention Sharing (AAS), a suite of three techniques meticulously designed to fully unleash the potential of attention sharing within the DiT. Furthermore, to broaden the applicability of our method, we present an efficient query, key, and value feature extraction algorithm, enabling our method to seamlessly incorporate external images as style references. Extensive experimental results validate that our method effectively enhances style consistency across generated images while maintaining precise text-to-image alignment.

Chongjian Yue, Lun Du, Qiang Fu et al.

With the development of temporal networks such as E-commerce networks and social networks, the issue of temporal link prediction has attracted increasing attention in recent years. The Temporal Link Prediction task of WSDM Cup 2022 expects a single model that can work well on two kinds of temporal graphs simultaneously, which have quite different characteristics and data properties, to predict whether a link of a given type will occur between two given nodes within a given time span. Our team, named as nothing here, regards this task as a link prediction task in heterogeneous temporal networks and proposes a generic model, i.e., Heterogeneous Temporal Graph Network (HTGN), to solve such temporal link prediction task with the unfixed time intervals and the diverse link types. That is, HTGN can adapt to the heterogeneity of links and the prediction with unfixed time intervals within an arbitrary given time period. To train the model, we design a Bi-Time-Window training strategy (BTW) which has two kinds of mini-batches from two kinds of time windows. As a result, for the final test, we achieved an AUC of 0.662482 on dataset A, an AUC of 0.906923 on dataset B, and won 2nd place with an Average T-scores of 0.628942.

Cagri Kilic, Bernardo Martinez R., Christopher A. Tatsch et al.

Plans for establishing a long-term human presence on the Moon will require substantial increases in robot autonomy and multi-robot coordination to support establishing a lunar outpost. To achieve these objectives, algorithm design choices for the software developments need to be tested and validated for expected scenarios such as autonomous in-situ resource utilization (ISRU), localization in challenging environments, and multi-robot coordination. However, real-world experiments are extremely challenging and limited for extraterrestrial environment. Also, realistic simulation demonstrations in these environments are still rare and demanded for initial algorithm testing capabilities. To help some of these needs, the NASA Centennial Challenges program established the Space Robotics Challenge Phase 2 (SRC2) which consist of virtual robotic systems in a realistic lunar simulation environment, where a group of mobile robots were tasked with reporting volatile locations within a global map, excavating and transporting these resources, and detecting and localizing a target of interest. The main goal of this article is to share our team's experiences on the design trade-offs to perform autonomous robotic operations in a virtual lunar environment and to share strategies to complete the mission requirements posed by NASA SRC2 competition during the qualification round. Of the 114 teams that registered for participation in the NASA SRC2, team Mountaineers finished as one of only six teams to receive the top qualification round prize.

Matteo De Petrillo, Jared Beard, Yu Gu et al.

We present a waypoint planning algorithm for an unmanned aerial vehicle (UAV) that is teamed with an unmanned ground vehicle (UGV) for the task of search and rescue in a subterranean environment. The UAV and UGV are teamed such that the localization of the UAV is conducted on the UGV via the multi-sensor fusion of a fish-eye camera, 3D LIDAR, ranging radio, and a laser altimeter. Likewise, the trajectory planning of the UAV is conducted on the UGV, which is assumed to have a 3D map of the environment (e.g., from Simultaneous Localization and Mapping). The goal of the planning algorithm is to satisfy the mission's exploration criteria while reducing the localization error of the UAV by evaluating the belief space for potential exploration routes. The presented algorithm is evaluated in a relevant simulation environment where the planning algorithm is shown to be effective at reducing the localization errors of the UAV.

Yu Gu, Yantong Wang, Tao Liu et al.

Emotion is well-recognized as a distinguished symbol of human beings, and it plays a crucial role in our daily lives. Existing vision-based or sensor-based solutions are either obstructive to use or rely on specialized hardware, hindering their applicability. This paper introduces EmoSense, a first-of-its-kind wireless emotion sensing system driven by computational intelligence. The basic methodology is to explore the physical expression of emotions from wireless channel response via data mining. The design and implementation of EmoSense {face} two major challenges: extracting physical expression from wireless channel data and recovering emotion from the corresponding physical expression. For the former, we present a Fresnel zone based theoretical model depicting the fingerprint of the physical expression on channel response. For the latter, we design an efficient computational intelligence driven mechanism to recognize emotion from the corresponding fingerprints. We prototyped EmoSense on the commodity WiFi infrastructure and compared it with main-stream sensor-based and vision-based approaches in the real-world scenario. The numerical study over $3360$ cases confirms that EmoSense achieves a comparable performance to the vision-based and sensor-based rivals under different scenarios. EmoSense only leverages the low-cost and prevalent WiFi infrastructures and thus constitutes a tempting solution for emotion sensing.

Yu Gu, Xiang Zhang, Zhi Liu et al.

Good quality sleep is essential for good health and sleep monitoring becomes a vital research topic. This paper provides a low cost, continuous and contactless WiFi-based vital signs (breathing and heart rate) monitoring method. In particular, we set up the antennas based on Fresnel diffraction model and signal propagation theory, which enhances the detection of weak breathing/heartbeat motion. We implement a prototype system using the off-shelf devices and a real-time processing system to monitor vital signs in real time. The experimental results indicate the accurate breathing rate and heart rate detection performance. To the best of our knowledge, this is the first work to use a pair of WiFi devices and omnidirectional antennas to achieve real-time individual breathing rate and heart rate monitoring in different sleeping postures.