Seonwook Park

Koki Nagano, Hongyu Liu, Seonwook Park et al.

We present DyaPlex, a streaming, full-duplex speech-and-motion model designed for dyadic interaction. To capture the continuous and reciprocal nature of human communication, this full-duplex capability empowers the agent to simultaneously perceive and generate both speech and physical motion in a streaming fashion. At its core, our method leverages the strong priors of a foundational full-duplex speech model and integrates a novel motion pathway, thereby achieving fully synchronized multi-modal interaction. Specifically, we design a dual-tower Transformer architecture that preserves the zero-shot conversational reasoning of a frozen base speech model while constructing a deeply coupled, streaming motion pathway. By introducing a unified dyadic token interleaving mechanism and guiding cross-attention via a time-aligned speech-motion RoPE, our model effectively aligns autoregressive motions with rich latent speech features. Trained on the 4,000-hour Seamless Interaction dataset, our model effectively captures cross-speaker dependencies and establishes new state-of-the-art performance across both monadic and dyadic human interaction benchmarks.

Chunggi Lee, Seonwook Park, Heon Song et al.

Annotating tens or hundreds of tiny objects in a given image is laborious yet crucial for a multitude of Computer Vision tasks. Such imagery typically contains objects from various categories, yet the multi-class interactive annotation setting for the detection task has thus far been unexplored. To address these needs, we propose a novel interactive annotation method for multiple instances of tiny objects from multiple classes, based on a few point-based user inputs. Our approach, C3Det, relates the full image context with annotator inputs in a local and global manner via late-fusion and feature-correlation, respectively. We perform experiments on the Tiny-DOTA and LCell datasets using both two-stage and one-stage object detection architectures to verify the efficacy of our approach. Our approach outperforms existing approaches in interactive annotation, achieving higher mAP with fewer clicks. Furthermore, we validate the annotation efficiency of our approach in a user study where it is shown to be 2.85x faster and yield only 0.36x task load (NASA-TLX, lower is better) compared to manual annotation. The code is available at https://github.com/ChungYi347/Interactive-Multi-Class-Tiny-Object-Detection.

Mingu Kang, Heon Song, Seonwook Park et al.

Computational pathology can lead to saving human lives, but models are annotation hungry and pathology images are notoriously expensive to annotate. Self-supervised learning has shown to be an effective method for utilizing unlabeled data, and its application to pathology could greatly benefit its downstream tasks. Yet, there are no principled studies that compare SSL methods and discuss how to adapt them for pathology. To address this need, we execute the largest-scale study of SSL pre-training on pathology image data, to date. Our study is conducted using 4 representative SSL methods on diverse downstream tasks. We establish that large-scale domain-aligned pre-training in pathology consistently out-performs ImageNet pre-training in standard SSL settings such as linear and fine-tuning evaluations, as well as in low-label regimes. Moreover, we propose a set of domain-specific techniques that we experimentally show leads to a performance boost. Lastly, for the first time, we apply SSL to the challenging task of nuclei instance segmentation and show large and consistent performance improvements under diverse settings.

Jeongun Ryu, Aaron Valero Puche, JaeWoong Shin et al.

Cell detection is a fundamental task in computational pathology that can be used for extracting high-level medical information from whole-slide images. For accurate cell detection, pathologists often zoom out to understand the tissue-level structures and zoom in to classify cells based on their morphology and the surrounding context. However, there is a lack of efforts to reflect such behaviors by pathologists in the cell detection models, mainly due to the lack of datasets containing both cell and tissue annotations with overlapping regions. To overcome this limitation, we propose and publicly release OCELOT, a dataset purposely dedicated to the study of cell-tissue relationships for cell detection in histopathology. OCELOT provides overlapping cell and tissue annotations on images acquired from multiple organs. Within this setting, we also propose multi-task learning approaches that benefit from learning both cell and tissue tasks simultaneously. When compared against a model trained only for the cell detection task, our proposed approaches improve cell detection performance on 3 datasets: proposed OCELOT, public TIGER, and internal CARP datasets. On the OCELOT test set in particular, we show up to 6.79 improvement in F1-score. We believe the contributions of this paper, including the release of the OCELOT dataset at https://lunit-io.github.io/research/publications/ocelot are a crucial starting point toward the important research direction of incorporating cell-tissue relationships in computation pathology.

Amrita Mazumdar, Seonwook Park, Rajarshi Roy et al.

Natural human conversation is full-duplex and audio-visual: people simultaneously speak and listen while continuously interpreting and producing nonverbal cues, such as nods, smiles, and gestures. To support successful human-agent interaction, agents must model full-duplex audiovisual conversation; however, existing full-duplex benchmarks evaluate only speech. In this work, we present VideoFDB, the first benchmark to evaluate full-duplex audio-visual-to-audio-visual (AV2AV) conversational agents. VideoFDB contributes (i) 237 dyadic clips spanning 11 nonverbal conversational dynamics from real-world video calls, (ii) a taxonomy separating perception from generation behaviors, and (iii) a rubric-based LM-as-judge evaluation framework with interpretable axes for assessing conversational quality with respect to nonverbal conversational dynamics. Across open- and closed-source vision-speech agents, we find systematic failure modes: captioning collapse and visual-stream ignorance, and we show that current systems exploit vision for explicit visual question answering but not for the streaming joint audiovisual grounding required in natural conversation. We further evaluate cascaded speech-to-avatar systems and find that their architecture fundamentally precludes the production of full-duplex nonverbal cues. As the first benchmark for full-duplex AV2AV interaction, VideoFDB establishes a foundation for systematic evaluation and, we hope, will accelerate the advancement and development of next-generation multimodal conversational agents.

Biagio Brattoli, Mohammad Mostafavi, Taebum Lee et al.

Despite advancements in methodologies, immunohistochemistry (IHC) remains the most utilized ancillary test for histopathologic and companion diagnostics in targeted therapies. However, objective IHC assessment poses challenges. Artificial intelligence (AI) has emerged as a potential solution, yet its development requires extensive training for each cancer and IHC type, limiting versatility. We developed a Universal IHC (UIHC) analyzer, an AI model for interpreting IHC images regardless of tumor or IHC types, using training datasets from various cancers stained for PD-L1 and/or HER2. This multi-cohort trained model outperforms conventional single-cohort models in interpreting unseen IHCs (Kappa score 0.578 vs. up to 0.509) and consistently shows superior performance across different positive staining cutoff values. Qualitative analysis reveals that UIHC effectively clusters patches based on expression levels. The UIHC model also quantitatively assesses c-MET expression with MET mutations, representing a significant advancement in AI application in the era of personalized medicine and accumulating novel biomarkers.

Kaiwen Jiang, Xueting Li, Seonwook Park et al.

Portrait animation has witnessed tremendous quality improvements thanks to recent advances in video diffusion models. However, these 2D methods often compromise 3D consistency and speed, limiting their applicability in real-world scenarios, such as digital twins or telepresence. In contrast, 3D-aware facial animation feedforward methods -- built upon explicit 3D representations, such as neural radiance fields or Gaussian splatting -- ensure 3D consistency and achieve faster inference speed, but come with inferior expression details. In this paper, we aim to combine their strengths by distilling knowledge from a 2D diffusion-based method into a feed-forward encoder, which instantly converts an in-the-wild single image into a 3D-consistent, fast yet expressive animatable representation. Our animation representation is decoupled from the face's 3D representation and learns motion implicitly from data, eliminating the dependency on pre-defined parametric models that often constrain animation capabilities. Unlike previous computationally intensive global fusion mechanisms (e.g., multiple attention layers) for fusing 3D structural and animation information, our design employs an efficient lightweight local fusion strategy to achieve high animation expressivity. As a result, our method runs at 107.31 FPS for animation and pose control while achieving comparable animation quality to the state-of-the-art, surpassing alternative designs that trade speed for quality or vice versa. Project website is https://research.nvidia.com/labs/amri/projects/instant4d

Rakshit Kothari, Shalini De Mello, Umar Iqbal et al.

A major challenge for physically unconstrained gaze estimation is acquiring training data with 3D gaze annotations for in-the-wild and outdoor scenarios. In contrast, videos of human interactions in unconstrained environments are abundantly available and can be much more easily annotated with frame-level activity labels. In this work, we tackle the previously unexplored problem of weakly-supervised gaze estimation from videos of human interactions. We leverage the insight that strong gaze-related geometric constraints exist when people perform the activity of "looking at each other" (LAEO). To acquire viable 3D gaze supervision from LAEO labels, we propose a training algorithm along with several novel loss functions especially designed for the task. With weak supervision from two large scale CMU-Panoptic and AVA-LAEO activity datasets, we show significant improvements in (a) the accuracy of semi-supervised gaze estimation and (b) cross-domain generalization on the state-of-the-art physically unconstrained in-the-wild Gaze360 gaze estimation benchmark. We open source our code at https://github.com/NVlabs/weakly-supervised-gaze.

Marcel Bühler, Seonwook Park, Shalini De Mello et al.

Accurately labeled real-world training data can be scarce, and hence recent works adapt, modify or generate images to boost target datasets. However, retaining relevant details from input data in the generated images is challenging and failure could be critical to the performance on the final task. In this work, we synthesize person-specific eye images that satisfy a given semantic segmentation mask (content), while following the style of a specified person from only a few reference images. We introduce two approaches, (a) one used to win the OpenEDS Synthetic Eye Generation Challenge at ICCV 2019, and (b) a principled approach to solving the problem involving simultaneous injection of style and content information at multiple scales. Our implementation is available at https://github.com/mcbuehler/Seg2Eye.

Seonwook Park, Shalini De Mello, Pavlo Molchanov et al.

Inter-personal anatomical differences limit the accuracy of person-independent gaze estimation networks. Yet there is a need to lower gaze errors further to enable applications requiring higher quality. Further gains can be achieved by personalizing gaze networks, ideally with few calibration samples. However, over-parameterized neural networks are not amenable to learning from few examples as they can quickly over-fit. We embrace these challenges and propose a novel framework for Few-shot Adaptive GaZE Estimation (FAZE) for learning person-specific gaze networks with very few (less than or equal to 9) calibration samples. FAZE learns a rotation-aware latent representation of gaze via a disentangling encoder-decoder architecture along with a highly adaptable gaze estimator trained using meta-learning. It is capable of adapting to any new person to yield significant performance gains with as few as 3 samples, yielding state-of-the-art performance of 3.18 degrees on GazeCapture, a 19% improvement over prior art. We open-source our code at https://github.com/NVlabs/few_shot_gaze

Haldun Balim, Seonwook Park, Xi Wang et al.

Despite the recent development of learning-based gaze estimation methods, most methods require one or more eye or face region crops as inputs and produce a gaze direction vector as output. Cropping results in a higher resolution in the eye regions and having fewer confounding factors (such as clothing and hair) is believed to benefit the final model performance. However, this eye/face patch cropping process is expensive, erroneous, and implementation-specific for different methods. In this paper, we propose a frame-to-gaze network that directly predicts both 3D gaze origin and 3D gaze direction from the raw frame out of the camera without any face or eye cropping. Our method demonstrates that direct gaze regression from the raw downscaled frame, from FHD/HD to VGA/HVGA resolution, is possible despite the challenges of having very few pixels in the eye region. The proposed method achieves comparable results to state-of-the-art methods in Point-of-Gaze (PoG) estimation on three public gaze datasets: GazeCapture, MPIIFaceGaze, and EVE, and generalizes well to extreme camera view changes.

Yufeng Zheng, Seonwook Park, Xucong Zhang et al.

Many computer vision tasks rely on labeled data. Rapid progress in generative modeling has led to the ability to synthesize photorealistic images. However, controlling specific aspects of the generation process such that the data can be used for supervision of downstream tasks remains challenging. In this paper we propose a novel generative model for images of faces, that is capable of producing high-quality images under fine-grained control over eye gaze and head orientation angles. This requires the disentangling of many appearance related factors including gaze and head orientation but also lighting, hue etc. We propose a novel architecture which learns to discover, disentangle and encode these extraneous variations in a self-learned manner. We further show that explicitly disentangling task-irrelevant factors results in more accurate modelling of gaze and head orientation. A novel evaluation scheme shows that our method improves upon the state-of-the-art in redirection accuracy and disentanglement between gaze direction and head orientation changes. Furthermore, we show that in the presence of limited amounts of real-world training data, our method allows for improvements in the downstream task of semi-supervised cross-dataset gaze estimation. Please check our project page at: https://ait.ethz.ch/projects/2020/STED-gaze/

Xucong Zhang, Seonwook Park, Thabo Beeler et al.

Gaze estimation is a fundamental task in many applications of computer vision, human computer interaction and robotics. Many state-of-the-art methods are trained and tested on custom datasets, making comparison across methods challenging. Furthermore, existing gaze estimation datasets have limited head pose and gaze variations, and the evaluations are conducted using different protocols and metrics. In this paper, we propose a new gaze estimation dataset called ETH-XGaze, consisting of over one million high-resolution images of varying gaze under extreme head poses. We collect this dataset from 110 participants with a custom hardware setup including 18 digital SLR cameras and adjustable illumination conditions, and a calibrated system to record ground truth gaze targets. We show that our dataset can significantly improve the robustness of gaze estimation methods across different head poses and gaze angles. Additionally, we define a standardized experimental protocol and evaluation metric on ETH-XGaze, to better unify gaze estimation research going forward. The dataset and benchmark website are available at https://ait.ethz.ch/projects/2020/ETH-XGaze

Seonwook Park, Emre Aksan, Xucong Zhang et al.

Estimating eye-gaze from images alone is a challenging task, in large parts due to un-observable person-specific factors. Achieving high accuracy typically requires labeled data from test users which may not be attainable in real applications. We observe that there exists a strong relationship between what users are looking at and the appearance of the user's eyes. In response to this understanding, we propose a novel dataset and accompanying method which aims to explicitly learn these semantic and temporal relationships. Our video dataset consists of time-synchronized screen recordings, user-facing camera views, and eye gaze data, which allows for new benchmarks in temporal gaze tracking as well as label-free refinement of gaze. Importantly, we demonstrate that the fusion of information from visual stimuli as well as eye images can lead towards achieving performance similar to literature-reported figures acquired through supervised personalization. Our final method yields significant performance improvements on our proposed EVE dataset, with up to a 28 percent improvement in Point-of-Gaze estimates (resulting in 2.49 degrees in angular error), paving the path towards high-accuracy screen-based eye tracking purely from webcam sensors. The dataset and reference source code are available at https://ait.ethz.ch/projects/2020/EVE

Anna Maria Feit, Lukas Vordemann, Seonwook Park et al.

This paper proposes an approach to detect information relevance during decision-making from eye movements in order to enable user interface adaptation. This is a challenging task because gaze behavior varies greatly across individual users and tasks and groundtruth data is difficult to obtain. Thus, prior work has mostly focused on simpler target-search tasks or on establishing general interest, where gaze behavior is less complex. From the literature, we identify six metrics that capture different aspects of the gaze behavior during decision-making and combine them in a voting scheme. We empirically show, that this accounts for the large variations in gaze behavior and out-performs standalone metrics. Importantly, it offers an intuitive way to control the amount of detected information, which is crucial for different UI adaptation schemes to succeed. We show the applicability of our approach by developing a room-search application that changes the visual saliency of content detected as relevant. In an empirical study, we show that it detects up to 97% of relevant elements with respect to user self-reporting, which allows us to meaningfully adapt the interface, as confirmed by participants. Our approach is fast, does not need any explicit user input and can be applied independent of task and user.

Seonwook Park, Adrian Spurr, Otmar Hilliges

Estimating human gaze from natural eye images only is a challenging task. Gaze direction can be defined by the pupil- and the eyeball center where the latter is unobservable in 2D images. Hence, achieving highly accurate gaze estimates is an ill-posed problem. In this paper, we introduce a novel deep neural network architecture specifically designed for the task of gaze estimation from single eye input. Instead of directly regressing two angles for the pitch and yaw of the eyeball, we regress to an intermediate pictorial representation which in turn simplifies the task of 3D gaze direction estimation. Our quantitative and qualitative results show that our approach achieves higher accuracies than the state-of-the-art and is robust to variation in gaze, head pose and image quality.

Seonwook Park, Xucong Zhang, Andreas Bulling et al.

Conventional feature-based and model-based gaze estimation methods have proven to perform well in settings with controlled illumination and specialized cameras. In unconstrained real-world settings, however, such methods are surpassed by recent appearance-based methods due to difficulties in modeling factors such as illumination changes and other visual artifacts. We present a novel learning-based method for eye region landmark localization that enables conventional methods to be competitive to latest appearance-based methods. Despite having been trained exclusively on synthetic data, our method exceeds the state of the art for iris localization and eye shape registration on real-world imagery. We then use the detected landmarks as input to iterative model-fitting and lightweight learning-based gaze estimation methods. Our approach outperforms existing model-fitting and appearance-based methods in the context of person-independent and personalized gaze estimation.

Adrian Spurr, Jie Song, Seonwook Park et al.

The human hand moves in complex and high-dimensional ways, making estimation of 3D hand pose configurations from images alone a challenging task. In this work we propose a method to learn a statistical hand model represented by a cross-modal trained latent space via a generative deep neural network. We derive an objective function from the variational lower bound of the VAE framework and jointly optimize the resulting cross-modal KL-divergence and the posterior reconstruction objective, naturally admitting a training regime that leads to a coherent latent space across multiple modalities such as RGB images, 2D keypoint detections or 3D hand configurations. Additionally, it grants a straightforward way of using semi-supervision. This latent space can be directly used to estimate 3D hand poses from RGB images, outperforming the state-of-the art in different settings. Furthermore, we show that our proposed method can be used without changes on depth images and performs comparably to specialized methods. Finally, the model is fully generative and can synthesize consistent pairs of hand configurations across modalities. We evaluate our method on both RGB and depth datasets and analyze the latent space qualitatively.

Seonwook Park, Christoph Gebhardt, Roman Rädle et al.

Developing cross-device multi-user interfaces (UIs) is a challenging problem. There are numerous ways in which content and interactivity can be distributed. However, good solutions must consider multiple users, their roles, their preferences and access rights, as well as device capabilities. Manual and rule-based solutions are tedious to create and do not scale to larger problems nor do they adapt to dynamic changes, such as users leaving or joining an activity. In this paper, we cast the problem of UI distribution as an assignment problem and propose to solve it using combinatorial optimization. We present a mixed integer programming formulation which allows real-time applications in dynamically changing collaborative settings. It optimizes the allocation of UI elements based on device capabilities, user roles, preferences, and access rights. We present a proof-of-concept designer-in-the-loop tool, allowing for quick solution exploration. Finally, we compare our approach to traditional paper prototyping in a lab study.