Tao Gao

Yuanbo Wen, Tao Gao, Jing Zhang et al.

The profound accumulation of precipitation during intense rainfall events can markedly degrade the quality of images, leading to the erosion of textural details. Despite the improvements observed in existing learning-based methods specialized for heavy rain removal, it is discerned that a significant proportion of these methods tend to overlook the precise reconstruction of the intricate details. In this work, we introduce a simple dual-stage progressive enhancement network, denoted as DPENet, aiming to achieve effective deraining while preserving the structural accuracy of rain-free images. This approach comprises two key modules, a rain streaks removal network (R$^2$Net) focusing on accurate rain removal, and a details reconstruction network (DRNet) designed to recover the textural details of rain-free images. Firstly, we introduce a dilated dense residual block (DDRB) within R$^2$Net, enabling the aggregation of high-level and low-level features. Secondly, an enhanced residual pixel-wise attention block (ERPAB) is integrated into DRNet to facilitate the incorporation of contextual information. To further enhance the fidelity of our approach, we employ a comprehensive loss function that accentuates both the marginal and regional accuracy of rain-free images. Extensive experiments conducted on publicly available benchmarks demonstrates the noteworthy efficiency and effectiveness of our proposed DPENet. The source code and pre-trained models are currently available at \url{https://github.com/chdwyb/DPENet}.

Tao Gao, Yuanbo Wen, Kaihao Zhang et al.

Rain-by-snow weather removal is a specialized task in weather-degraded image restoration aiming to eliminate coexisting rain streaks and snow particles. In this paper, we propose RSFormer, an efficient and effective Transformer that addresses this challenge. Initially, we explore the proximity of convolution networks (ConvNets) and vision Transformers (ViTs) in hierarchical architectures and experimentally find they perform approximately at intra-stage feature learning. On this basis, we utilize a Transformer-like convolution block (TCB) that replaces the computationally expensive self-attention while preserving attention characteristics for adapting to input content. We also demonstrate that cross-stage progression is critical for performance improvement, and propose a global-local self-attention sampling mechanism (GLASM) that down-/up-samples features while capturing both global and local dependencies. Finally, we synthesize two novel rain-by-snow datasets, RSCityScape and RS100K, to evaluate our proposed RSFormer. Extensive experiments verify that RSFormer achieves the best trade-off between performance and time-consumption compared to other restoration methods. For instance, it outperforms Restormer with a 1.53% reduction in the number of parameters and a 15.6% reduction in inference time. Datasets, source code and pre-trained models are available at \url{https://github.com/chdwyb/RSFormer}.

Yuanbo Wen, Tao Gao, Ziqi Li et al.

Eliminating the rain degradation in stereo images poses a formidable challenge, which necessitates the efficient exploitation of mutual information present between the dual views. To this end, we devise MQINet, which employs multi-dimension queries and interactions for stereo image deraining. More specifically, our approach incorporates a context-aware dimension-wise queried block (CDQB). This module leverages dimension-wise queries that are independent of the input features and employs global context-aware attention (GCA) to capture essential features while avoiding the entanglement of redundant or irrelevant information. Meanwhile, we introduce an intra-view physics-aware attention (IPA) based on the inverse physical model of rainy images. IPA extracts shallow features that are sensitive to the physics of rain degradation, facilitating the reduction of rain-related artifacts during the early learning period. Furthermore, we integrate a cross-view multi-dimension interacting attention mechanism (CMIA) to foster comprehensive feature interaction between the two views across multiple dimensions. Extensive experimental evaluations demonstrate the superiority of our model over EPRRNet and StereoIRR, achieving respective improvements of 4.18 dB and 0.45 dB in PSNR. Code and models are available at \url{https://github.com/chdwyb/MQINet}.

Zuoyin Tang, Jianhua He, Dashuai Pei et al.

Handling long tail corner cases is a major challenge faced by autonomous vehicles (AVs). While large language models (LLMs) hold great potentials to handle the corner cases with excellent generalization and explanation capabilities and received increasing research interest on application to autonomous driving, there are still technical barriers to be tackled, such as strict model performance and huge computing resource requirements of LLMs. In this paper, we investigate a new approach of applying remote or edge LLMs to support autonomous driving. A key issue for such LLM assisted driving system is the assessment of LLMs on their understanding of driving theory and skills, ensuring they are qualified to undertake safety critical driving assistance tasks for CAVs. We design and run driving theory tests for several proprietary LLM models (OpenAI GPT models, Baidu Ernie and Ali QWen) and open-source LLM models (Tsinghua MiniCPM-2B and MiniCPM-Llama3-V2.5) with more than 500 multiple-choices theory test questions. Model accuracy, cost and processing latency are measured from the experiments. Experiment results show that while model GPT-4 passes the test with improved domain knowledge and Ernie has an accuracy of 85% (just below the 86% passing threshold), other LLM models including GPT-3.5 fail the test. For the test questions with images, the multimodal model GPT4-o has an excellent accuracy result of 96%, and the MiniCPM-Llama3-V2.5 achieves an accuracy of 76%. While GPT-4 holds stronger potential for CAV driving assistance applications, the cost of using model GPT4 is much higher, almost 50 times of that of using GPT3.5. The results can help make decision on the use of the existing LLMs for CAV applications and balancing on the model performance and cost.

Yuanbo Wen, Tao Gao, Ting Chen

Existing unpaired image deraining approaches face challenges in accurately capture the distinguishing characteristics between the rainy and clean domains, resulting in residual degradation and color distortion within the reconstructed images. To this end, we propose an energy-informed diffusion model for unpaired photo-realistic image deraining (UPID-EDM). Initially, we delve into the intricate visual-language priors embedded within the contrastive language-image pre-training model (CLIP), and demonstrate that the CLIP priors aid in the discrimination of rainy and clean images. Furthermore, we introduce a dual-consistent energy function (DEF) that retains the rain-irrelevant characteristics while eliminating the rain-relevant features. This energy function is trained by the non-corresponding rainy and clean images. In addition, we employ the rain-relevance discarding energy function (RDEF) and the rain-irrelevance preserving energy function (RPEF) to direct the reverse sampling procedure of a pre-trained diffusion model, effectively removing the rain streaks while preserving the image contents. Extensive experiments demonstrate that our energy-informed model surpasses the existing unpaired learning approaches in terms of both supervised and no-reference metrics.

Yufei Wang, Bo Li, Ge Zhang et al.

Existing deep learning-based depth completion methods generally employ massive stacked layers to predict the dense depth map from sparse input data. Although such approaches greatly advance this task, their accompanied huge computational complexity hinders their practical applications. To accomplish depth completion more efficiently, we propose a novel lightweight deep network framework, the Long-short Range Recurrent Updating (LRRU) network. Without learning complex feature representations, LRRU first roughly fills the sparse input to obtain an initial dense depth map, and then iteratively updates it through learned spatially-variant kernels. Our iterative update process is content-adaptive and highly flexible, where the kernel weights are learned by jointly considering the guidance RGB images and the depth map to be updated, and large-to-small kernel scopes are dynamically adjusted to capture long-to-short range dependencies. Our initial depth map has coarse but complete scene depth information, which helps relieve the burden of directly regressing the dense depth from sparse ones, while our proposed method can effectively refine it to an accurate depth map with less learnable parameters and inference time. Experimental results demonstrate that our proposed LRRU variants achieve state-of-the-art performance across different parameter regimes. In particular, the LRRU-Base model outperforms competing approaches on the NYUv2 dataset, and ranks 1st on the KITTI depth completion benchmark at the time of submission. Project page: https://npucvr.github.io/LRRU/.

Yuanbo Wen, Tao Gao, Ziqi Li et al.

Haze obscures remote sensing images, hindering valuable information extraction. To this end, we propose RSHazeNet, an encoder-minimal and decoder-minimal framework for efficient remote sensing image dehazing. Specifically, regarding the process of merging features within the same level, we develop an innovative module called intra-level transposed fusion module (ITFM). This module employs adaptive transposed self-attention to capture comprehensive context-aware information, facilitating the robust context-aware feature fusion. Meanwhile, we present a cross-level multi-view interaction module (CMIM) to enable effective interactions between features from various levels, mitigating the loss of information due to the repeated sampling operations. In addition, we propose a multi-view progressive extraction block (MPEB) that partitions the features into four distinct components and employs convolution with varying kernel sizes, groups, and dilation factors to facilitate view-progressive feature learning. Extensive experiments demonstrate the superiority of our proposed RSHazeNet. We release the source code and all pre-trained models at \url{https://github.com/chdwyb/RSHazeNet}.

Jiale Wei, Xiang Ying, Tao Gao et al.

Human interaction with the external world fundamentally involves the exchange of personal memory, whether with other individuals, websites, applications, or, in the future, AI agents. A significant portion of this interaction is redundant, requiring users to repeatedly provide the same information across different contexts. Existing solutions, such as browser-stored credentials, autofill mechanisms, and unified authentication systems, have aimed to mitigate this redundancy by serving as intermediaries that store and retrieve commonly used user data. The advent of large language models (LLMs) presents an opportunity to redefine memory management through an AI-native paradigm: SECOND ME. SECOND ME acts as an intelligent, persistent memory offload system that retains, organizes, and dynamically utilizes user-specific knowledge. By serving as an intermediary in user interactions, it can autonomously generate context-aware responses, prefill required information, and facilitate seamless communication with external systems, significantly reducing cognitive load and interaction friction. Unlike traditional memory storage solutions, SECOND ME extends beyond static data retention by leveraging LLM-based memory parameterization. This enables structured organization, contextual reasoning, and adaptive knowledge retrieval, facilitating a more systematic and intelligent approach to memory management. As AI-driven personal agents like SECOND ME become increasingly integrated into digital ecosystems, SECOND ME further represents a critical step toward augmenting human-world interaction with persistent, contextually aware, and self-optimizing memory systems. We have open-sourced the fully localizable deployment system at GitHub: https://github.com/Mindverse/Second-Me.

Jinxiang Lai, Zexin Lu, Jiajun He et al.

Visual content creation tasks demand a nuanced understanding of design conventions and creative workflows-capabilities challenging for general models, while workflow-based agents lack specialized knowledge for autonomous creative planning. To overcome these challenges, we propose VisionCreator, a native visual-generation agentic model that unifies Understanding, Thinking, Planning, and Creation (UTPC) capabilities within an end-to-end learnable framework. Our work introduces four key contributions: (i) VisGenData-4k and its construction methodology using metacognition-based VisionAgent to generate high-quality creation trajectories with explicit UTPC structures; (ii) The VisionCreator agentic model, optimized through Progressive Specialization Training (PST) and Virtual Reinforcement Learning (VRL) within a high-fidelity simulated environment, enabling stable and efficient acquisition of UTPC capabilities for complex creation tasks; (iii) VisGenBench, a comprehensive benchmark featuring 1.2k test samples across diverse scenarios for standardized evaluation of multi-step visual creation capabilities; (iv) Remarkably, our VisionCreator-8B/32B models demonstrate superior performance over larger closed-source models across multiple evaluation dimensions. Overall, this work provides a foundation for future research in visual-generation agentic systems.

Yuanbo Wen, Tao Gao, Ziqi Li et al.

Existing approaches for all-in-one weather-degraded image restoration suffer from inefficiencies in leveraging degradation-aware priors, resulting in sub-optimal performance in adapting to different weather conditions. To this end, we develop an adaptive degradation-aware self-prompting model (ADSM) for all-in-one weather-degraded image restoration. Specifically, our model employs the contrastive language-image pre-training model (CLIP) to facilitate the training of our proposed latent prompt generators (LPGs), which represent three types of latent prompts to characterize the degradation type, degradation property and image caption. Moreover, we integrate the acquired degradation-aware prompts into the time embedding of diffusion model to improve degradation perception. Meanwhile, we employ the latent caption prompt to guide the reverse sampling process using the cross-attention mechanism, thereby guiding the accurate image reconstruction. Furthermore, to accelerate the reverse sampling procedure of diffusion model and address the limitations of frequency perception, we introduce a wavelet-oriented noise estimating network (WNE-Net). Extensive experiments conducted on eight publicly available datasets demonstrate the effectiveness of our proposed approach in both task-specific and all-in-one applications.

Minglu Zhao, Dehong Xu, Tao Gao

Attention is a cornerstone of human cognition that facilitates the efficient extraction of information in everyday life. Recent developments in artificial intelligence like the Transformer architecture also incorporate the idea of attention in model designs. However, despite the shared fundamental principle of selectively attending to information, human attention and the Transformer model display notable differences, particularly in their capacity constraints, attention pathways, and intentional mechanisms. Our review aims to provide a comparative analysis of these mechanisms from a cognitive-functional perspective, thereby shedding light on several open research questions. The exploration encourages interdisciplinary efforts to derive insights from human attention mechanisms in the pursuit of developing more generalized artificial intelligence.

Yehui Tang, Yichun Yin, Yaoyuan Wang et al.

Sparse large language models (LLMs) with Mixture of Experts (MoE) and close to a trillion parameters are dominating the realm of most capable language models. However, the massive model scale poses significant challenges for the underlying software and hardware systems. In this paper, we aim to uncover a recipe to harness such scale on Ascend NPUs. The key goals are better usage of the computing resources under the dynamic sparse model structures and materializing the expected performance gain on the actual hardware. To select model configurations suitable for Ascend NPUs without repeatedly running the expensive experiments, we leverage simulation to compare the trade-off of various model hyperparameters. This study led to Pangu Ultra MoE, a sparse LLM with 718 billion parameters, and we conducted experiments on the model to verify the simulation results. On the system side, we dig into Expert Parallelism to optimize the communication between NPU devices to reduce the synchronization overhead. We also optimize the memory efficiency within the devices to further reduce the parameter and activation management overhead. In the end, we achieve an MFU of 30.0% when training Pangu Ultra MoE, with performance comparable to that of DeepSeek R1, on 6K Ascend NPUs, and demonstrate that the Ascend system is capable of harnessing all the training stages of the state-of-the-art language models. Extensive experiments indicate that our recipe can lead to efficient training of large-scale sparse language models with MoE. We also study the behaviors of such models for future reference.

Jun Yu Chen, Tao Gao

We present APT, an advanced Large Language Model (LLM)-driven framework that enables autonomous agents to construct complex and creative structures within the Minecraft environment. Unlike previous approaches that primarily concentrate on skill-based open-world tasks or rely on image-based diffusion models for generating voxel-based structures, our method leverages the intrinsic spatial reasoning capabilities of LLMs. By employing chain-of-thought decomposition along with multimodal inputs, the framework generates detailed architectural layouts and blueprints that the agent can execute under zero-shot or few-shot learning scenarios. Our agent incorporates both memory and reflection modules to facilitate lifelong learning, adaptive refinement, and error correction throughout the building process. To rigorously evaluate the agent's performance in this emerging research area, we introduce a comprehensive benchmark consisting of diverse construction tasks designed to test creativity, spatial reasoning, adherence to in-game rules, and the effective integration of multimodal instructions. Experimental results using various GPT-based LLM backends and agent configurations demonstrate the agent's capacity to accurately interpret extensive instructions involving numerous items, their positions, and orientations. The agent successfully produces complex structures complete with internal functionalities such as Redstone-powered systems. A/B testing indicates that the inclusion of a memory module leads to a significant increase in performance, emphasizing its role in enabling continuous learning and the reuse of accumulated experience. Additionally, the agent's unexpected emergence of scaffolding behavior highlights the potential of future LLM-driven agents to utilize subroutine planning and leverage the emergence ability of LLMs to autonomously develop human-like problem-solving techniques.

Qian Long, Ruoyan Li, Minglu Zhao et al.

A major challenge for Multi-Agent Systems is enabling agents to adapt dynamically to diverse environments in which opponents and teammates may continually change. Agents trained using conventional methods tend to excel only within the confines of their training cohorts; their performance drops significantly when confronting unfamiliar agents. To address this shortcoming, we introduce Inverse Attention Agents that adopt concepts from the Theory of Mind (ToM) implemented algorithmically using an attention mechanism trained in an end-to-end manner. Crucial to determining the final actions of these agents, the weights in their attention model explicitly represent attention to different goals. We furthermore propose an inverse attention network that deduces the ToM of agents based on observations and prior actions. The network infers the attentional states of other agents, thereby refining the attention weights to adjust the agent's final action. We conduct experiments in a continuous environment, tackling demanding tasks encompassing cooperation, competition, and a blend of both. They demonstrate that the inverse attention network successfully infers the attention of other agents, and that this information improves agent performance. Additional human experiments show that, compared to baseline agent models, our inverse attention agents exhibit superior cooperation with humans and better emulate human behaviors.

Yingdong Qian, Marta Kryven, Tao Gao et al.

How to build AI that understands human intentions, and uses this knowledge to collaborate with people? We describe a computational framework for evaluating models of goal inference in the domain of 3D motor actions, which receives as input the 3D coordinates of an agent's body, and of possible targets, to produce a continuously updated inference of the intended target. We evaluate our framework in three behavioural experiments using a novel Target Reaching Task, in which human observers infer intentions of actors reaching for targets among distracts. We describe Generative Body Kinematics model, which predicts human intention inference in this domain using Bayesian inverse planning and inverse body kinematics. We compare our model to three heuristics, which formalize the principle of least effort using simple assumptions about the actor's constraints, without the use of inverse planning. Despite being more computationally costly, the Generative Body Kinematics model outperforms the heuristics in certain scenarios, such as environments with obstacles, and at the beginning of reaching actions while the actor is relatively far from the intended target. The heuristics make increasingly accurate predictions during later stages of reaching actions, such as, when the intended target is close, and can be inferred by extrapolating the wrist trajectory. Our results identify contexts in which inverse body kinematics is useful for intention inference. We show that human observers indeed rely on inverse body kinematics in such scenarios, suggesting that modeling body kinematic can improve performance of inference algorithms.

Shuwen Qiu, Sirui Xie, Lifeng Fan et al.

Humans communicate with graphical sketches apart from symbolic languages. Primarily focusing on the latter, recent studies of emergent communication overlook the sketches; they do not account for the evolution process through which symbolic sign systems emerge in the trade-off between iconicity and symbolicity. In this work, we take the very first step to model and simulate this process via two neural agents playing a visual communication game; the sender communicates with the receiver by sketching on a canvas. We devise a novel reinforcement learning method such that agents are evolved jointly towards successful communication and abstract graphical conventions. To inspect the emerged conventions, we define three fundamental properties -- iconicity, symbolicity, and semanticity -- and design evaluation methods accordingly. Our experimental results under different controls are consistent with the observation in studies of human graphical conventions. Of note, we find that evolved sketches can preserve the continuum of semantics under proper environmental pressures. More interestingly, co-evolved agents can switch between conventionalized and iconic communication based on their familiarity with referents. We hope the present research can pave the path for studying emergent communication with the modality of sketches.

Yixin Chen, Qing Li, Deqian Kong et al.

We study the understanding of embodied reference: One agent uses both language and gesture to refer to an object to another agent in a shared physical environment. Of note, this new visual task requires understanding multimodal cues with perspective-taking to identify which object is being referred to. To tackle this problem, we introduce YouRefIt, a new crowd-sourced dataset of embodied reference collected in various physical scenes; the dataset contains 4,195 unique reference clips in 432 indoor scenes. To the best of our knowledge, this is the first embodied reference dataset that allows us to study referring expressions in daily physical scenes to understand referential behavior, human communication, and human-robot interaction. We further devise two benchmarks for image-based and video-based embodied reference understanding. Comprehensive baselines and extensive experiments provide the very first result of machine perception on how the referring expressions and gestures affect the embodied reference understanding. Our results provide essential evidence that gestural cues are as critical as language cues in understanding the embodied reference.

Stephanie Stacy, Chenfei Li, Minglu Zhao et al.

Communication is highly overloaded. Despite this, even young children are good at leveraging context to understand ambiguous signals. We propose a computational account of overloaded signaling from a shared agency perspective which we call the Imagined We for Communication. Under this framework, communication helps cooperators coordinate their perspectives, allowing them to act together to achieve shared goals. We assume agents are rational cooperators, which puts constraints on how signals can be sent and interpreted. We implement this model in a set of simulations demonstrating this model's success under increasing ambiguity as well as increasing layers of reasoning. Our model is capable of improving performance with deeper recursive reasoning; however, it outperforms comparison baselines at even the shallowest level, highlighting how shared knowledge and cooperative logic can do much of the heavy-lifting in language.

Kaiwen Jiang, Stephanie Stacy, Chuyu Wei et al.

The simple gesture of pointing can greatly augment ones ability to comprehend states of the world based on observations. It triggers additional inferences relevant to ones task at hand. We model an agents update to its belief of the world based on individual observations using a partially observable Markov decision process (POMDP), a mainstream artificial intelligence (AI) model of how to act rationally according to beliefs formed through observation. On top of that, we model pointing as a communicative act between agents who have a mutual understanding that the pointed observation must be relevant and interpretable. Our model measures relevance by defining a Smithian Value of Information (SVI) as the utility improvement of the POMDP agent before and after receiving the pointing. We model that agents calculate SVI by using the cognitive theory of Smithian helping as a principle of coordinating separate beliefs for action prediction and action evaluation. We then import SVI into rational speech act (RSA) as the utility function of an utterance. These lead us to a pragmatic model of pointing allowing for contextually flexible interpretations. We demonstrate the power of our Smithian pointing model by extending the Wumpus world, a classic AI task where a hunter hunts a monster with only partial observability of the world. We add another agent as a guide who can only help by marking an observation already perceived by the hunter with a pointing or not, without providing new observations or offering any instrumental help. Our results show that this severely limited and overloaded communication nevertheless significantly improves the hunters performance. The advantage of pointing is indeed due to a computation of relevance based on Smithian helping, as it disappears completely when the task is too difficult or too easy for the guide to help.

Lifeng Fan, Shuwen Qiu, Zilong Zheng et al.

Humans possess a unique social cognition capability; nonverbal communication can convey rich social information among agents. In contrast, such crucial social characteristics are mostly missing in the existing scene understanding literature. In this paper, we incorporate different nonverbal communication cues (e.g., gaze, human poses, and gestures) to represent, model, learn, and infer agents' mental states from pure visual inputs. Crucially, such a mental representation takes the agent's belief into account so that it represents what the true world state is and infers the beliefs in each agent's mental state, which may differ from the true world states. By aggregating different beliefs and true world states, our model essentially forms "five minds" during the interactions between two agents. This "five minds" model differs from prior works that infer beliefs in an infinite recursion; instead, agents' beliefs are converged into a "common mind". Based on this representation, we further devise a hierarchical energy-based model that jointly tracks and predicts all five minds. From this new perspective, a social event is interpreted by a series of nonverbal communication and belief dynamics, which transcends the classic keyframe video summary. In the experiments, we demonstrate that using such a social account provides a better video summary on videos with rich social interactions compared with state-of-the-art keyframe video summary methods.

Tao Yuan, Hangxin Liu, Lifeng Fan et al.

Aiming to understand how human (false-)belief--a core socio-cognitive ability--would affect human interactions with robots, this paper proposes to adopt a graphical model to unify the representation of object states, robot knowledge, and human (false-)beliefs. Specifically, a parse graph (pg) is learned from a single-view spatiotemporal parsing by aggregating various object states along the time; such a learned representation is accumulated as the robot's knowledge. An inference algorithm is derived to fuse individual pg from all robots across multi-views into a joint pg, which affords more effective reasoning and inference capability to overcome the errors originated from a single view. In the experiments, through the joint inference over pg-s, the system correctly recognizes human (false-)belief in various settings and achieves better cross-view accuracy on a challenging small object tracking dataset.

Yixin Zhu, Tao Gao, Lifeng Fan et al.

Recent progress in deep learning is essentially based on a "big data for small tasks" paradigm, under which massive amounts of data are used to train a classifier for a single narrow task. In this paper, we call for a shift that flips this paradigm upside down. Specifically, we propose a "small data for big tasks" paradigm, wherein a single artificial intelligence (AI) system is challenged to develop "common sense", enabling it to solve a wide range of tasks with little training data. We illustrate the potential power of this new paradigm by reviewing models of common sense that synthesize recent breakthroughs in both machine and human vision. We identify functionality, physics, intent, causality, and utility (FPICU) as the five core domains of cognitive AI with humanlike common sense. When taken as a unified concept, FPICU is concerned with the questions of "why" and "how", beyond the dominant "what" and "where" framework for understanding vision. They are invisible in terms of pixels but nevertheless drive the creation, maintenance, and development of visual scenes. We therefore coin them the "dark matter" of vision. Just as our universe cannot be understood by merely studying observable matter, we argue that vision cannot be understood without studying FPICU. We demonstrate the power of this perspective to develop cognitive AI systems with humanlike common sense by showing how to observe and apply FPICU with little training data to solve a wide range of challenging tasks, including tool use, planning, utility inference, and social learning. In summary, we argue that the next generation of AI must embrace "dark" humanlike common sense for solving novel tasks.

Daniel Harari, Tao Gao, Nancy Kanwisher et al.

Humans are remarkably adept at interpreting the gaze direction of other individuals in their surroundings. This skill is at the core of the ability to engage in joint visual attention, which is essential for establishing social interactions. How accurate are humans in determining the gaze direction of others in lifelike scenes, when they can move their heads and eyes freely, and what are the sources of information for the underlying perceptual processes? These questions pose a challenge from both empirical and computational perspectives, due to the complexity of the visual input in real-life situations. Here we measure empirically human accuracy in perceiving the gaze direction of others in lifelike scenes, and study computationally the sources of information and representations underlying this cognitive capacity. We show that humans perform better in face-to-face conditions compared with recorded conditions, and that this advantage is not due to the availability of input dynamics. We further show that humans are still performing well when only the eyes-region is visible, rather than the whole face. We develop a computational model, which replicates the pattern of human performance, including the finding that the eyes-region contains on its own, the required information for estimating both head orientation and direction of gaze. Consistent with neurophysiological findings on task-specific face regions in the brain, the learned computational representations reproduce perceptual effects such as the Wollaston illusion, when trained to estimate direction of gaze, but not when trained to recognize objects or faces.

Tao Gao, Daniel Harari, Joshua Tenenbaum et al.

Joint attention is a core, early-developing form of social interaction. It is based on our ability to discriminate the third party objects that other people are looking at. While it has been shown that people can accurately determine whether another person is looking directly at them versus away, little is known about human ability to discriminate a third person gaze directed towards objects that are further away, especially in unconstraint cases where the looker can move her head and eyes freely. In this paper we address this question by jointly exploring human psychophysics and a cognitively motivated computer vision model, which can detect the 3D direction of gaze from 2D face images. The synthesis of behavioral study and computer vision yields several interesting discoveries. (1) Human accuracy of discriminating targets 8°-10° of visual angle apart is around 40% in a free looking gaze task; (2) The ability to interpret gaze of different lookers vary dramatically; (3) This variance can be captured by the computational model; (4) Human outperforms the current model significantly. These results collectively show that the acuity of human joint attention is indeed highly impressive, given the computational challenge of the natural looking task. Moreover, the gap between human and model performance, as well as the variability of gaze interpretation across different lookers, require further understanding of the underlying mechanisms utilized by humans for this challenging task.