arXiv subjects

Jingjing Qian

Publications and source records attributed to Jingjing Qian.

10 recordsLinked to original sources

arXiv2026

AVA-VLA: Improving Vision-Language-Action models with Active Visual Attention

Vision-Language-Action (VLA) models have shown remarkable progress in embodied tasks recently, but most methods process visual observations independently at each timestep. This history-agnostic design treats robot manipulation as a Markov Decision Process, even though real-world robotic control is inherently partially observable and requires reasoning over past interactions. To address this mismatch, we reformulate VLA policy learning from a Partially Observable Markov Decision Process perspective and propose AVA-VLA, a framework that conditions action generation on a recurrent state that serves as a neural approximation to the agent's belief over task history. Built on this recurrent state, we introduce Active Visual Attention (AVA), which dynamically reweights visual tokens in the current observation to focus on regions most relevant given both the instruction and execution history. Extensive experiments show that AVA-VLA achieves state-of-the-art performance on standard robotic benchmarks, including LIBERO and CALVIN, and transfers effectively to real-world dual-arm manipulation tasks. These results demonstrate the effectiveness of temporally grounded active visual processing for improving VLA performance in robotic sequential decision-making. The project page is available at https://liauto-dsr.github.io/AVA-VLA-Page.

cs.LG↗

arXiv2026

ELAN4D: Embodiment-Centric 4D Supervision for Vision-Language-Action Models via Plug-and-Play Adaptation

Vision-Language-Action (VLA) models have shown promise for robotic manipulation, yet most existing policies operate reactively by directly regressing actions from current observations, without explicitly modeling future dynamics. This limits their ability to generalize under out-of-distribution perturbations. To address this issue, we propose ELAN4D, an embodiment-centric, 4D-aware training framework that enhances VLA policies with future robot keypoint tracks as predictive spatio-temporal supervision. Using only forward kinematics from proprioceptive states, we derive 3D displacement tracks of robot keypoints, such as joints and the end-effector, with negligible preprocess cost. These tracks provide metric and compact supervision without requiring external trackers or reconstruction. A plug-and-play auxiliary branch with a lightweight track decoder injects this 4D signal into the action expert while preserving the pretrained vision-language backbone through gradient isolation. The track decoder is discarded during inference, leaving the base policy interface unchanged. Extensive experiments on LIBERO, LIBERO-Plus, RoboTwin2.0 and real-world manipulation tasks demonstrate that ELAN4D consistently improves over strong VLA baselines, achieving the best overall performance and substantial gains under out-of-distribution perturbations, including camera, background, and layout shifts. These results highlight the effectiveness of embodiment-centric 4D supervision for building more robust and generalizable manipulation policies.

cs.RO↗

arXiv2026

ESCAPE: Episodic Spatial Memory and Adaptive Execution Policy for Long-Horizon Mobile Manipulation

Coordinating navigation and manipulation with robust performance is essential for embodied AI in complex indoor environments. However, as tasks extend over long horizons, existing methods often struggle due to catastrophic forgetting, spatial inconsistency, and rigid execution. To address these issues, we propose ESCAPE (Episodic Spatial Memory Coupled with an Adaptive Policy for Execution), operating through a tightly coupled perception-grounding-execution workflow. For robust perception, ESCAPE features a Spatio-Temporal Fusion Mapping module to autoregressively construct a depth-free, persistent 3D spatial memory, alongside a Memory-Driven Target Grounding module for precise interaction mask generation. To achieve flexible action, our Adaptive Execution Policy dynamically orchestrates proactive global navigation and reactive local manipulation to seize opportunistic targets. ESCAPE achieves state-of-the-art performance on the ALFRED benchmark, reaching 65.09% and 60.79% success rates in test seen and unseen environments with step-by-step instructions. By reducing redundant exploration, our ESCAPE attains substantial improvements in path-length-weighted metrics and maintains robust performance (61.24% / 56.04%) even without detailed guidance for long-horizon tasks.

cs.CV↗

arXiv2026

JoyAI-LLM Flash: Advancing Mid-Scale LLMs with Token Efficiency

We introduce JoyAI-LLM Flash, an efficient Mixture-of-Experts (MoE) language model designed to redefine the trade-off between strong performance and token efficiency in the sub-50B parameter regime. JoyAI-LLM Flash is pretrained on a massive corpus of 20 trillion tokens and further optimized through a rigorous post-training pipeline, including supervised fine-tuning (SFT), Direct Preference Optimization (DPO), and large-scale reinforcement learning (RL) across diverse environments. To improve token efficiency, JoyAI-LLM Flash strategically balances \emph{thinking} and \emph{non-thinking} cognitive modes and introduces FiberPO, a novel RL algorithm inspired by fibration theory that decomposes trust-region maintenance into global and local components, providing unified multi-scale stability control for LLM policy optimization. To enhance architectural sparsity, the model comprises 48B total parameters while activating only 2.7B parameters per forward pass, achieving a substantially higher sparsity ratio than contemporary industry leading models of comparable scale. To further improve inference throughput, we adopt a joint training-inference co-design that incorporates dense Multi-Token Prediction (MTP) and Quantization-Aware Training (QAT). We release the checkpoints for both JoyAI-LLM-48B-A3B Base and its post-trained variants on Hugging Face to support the open-source community.

cs.CL↗

arXiv2026

GeoPredict: Leveraging Predictive Kinematics and 3D Gaussian Geometry for Precise VLA Manipulation

Vision-Language-Action (VLA) models achieve strong generalization in robotic manipulation but remain largely reactive and 2D-centric, making them unreliable in tasks that require precise 3D reasoning. We propose GeoPredict, a geometry-aware VLA framework that augments a continuous-action policy with predictive kinematic and geometric priors. GeoPredict introduces a trajectory-level module that encodes motion history and predicts multi-step 3D keypoint trajectories of robot arms, and a predictive 3D Gaussian geometry module that forecasts workspace geometry with track-guided refinement along future keypoint trajectories. These predictive modules serve exclusively as training-time supervision through depth-based rendering, while inference requires only lightweight additional query tokens without invoking any 3D decoding. Experiments on RoboCasa Human-50, LIBERO, and real-world manipulation tasks show that GeoPredict consistently outperforms strong VLA baselines, especially in geometry-intensive and spatially demanding scenarios.

cs.CV↗

arXiv2026

Proteus-ID: ID-Consistent and Motion-Coherent Video Customization

Video identity customization seeks to synthesize realistic, temporally coherent videos of a specific subject, given a single reference image and a text prompt. This task presents two core challenges: (1) maintaining identity consistency while aligning with the described appearance and actions, and (2) generating natural, fluid motion without unrealistic stiffness. To address these challenges, we introduce Proteus-ID, a novel diffusion-based framework for identity-consistent and motion-coherent video customization. First, we propose a Multimodal Identity Fusion (MIF) module that unifies visual and textual cues into a joint identity representation using a Q-Former, providing coherent guidance to the diffusion model and eliminating modality imbalance. Second, we present a Time-Aware Identity Injection (TAII) mechanism that dynamically modulates identity conditioning across denoising steps, improving fine-detail reconstruction. Third, we propose Adaptive Motion Learning (AML), a self-supervised strategy that reweights the training loss based on optical-flow-derived motion heatmaps, enhancing motion realism without requiring additional inputs. To support this task, we construct Proteus-Bench, a high-quality dataset comprising 200K curated clips for training and 150 individuals from diverse professions and ethnicities for evaluation. Extensive experiments demonstrate that Proteus-ID outperforms prior methods in identity preservation, text alignment, and motion quality, establishing a new benchmark for video identity customization. Codes and data are publicly available at https://grenoble-zhang.github.io/Proteus-ID/.

cs.CV↗

arXiv2025

Rate-Distortion-Perception Tradeoff for Gaussian Vector Sources

This paper studies the rate-distortion-perception (RDP) tradeoff for a Gaussian vector source coding problem where the goal is to compress the multi-component source subject to distortion and perception constraints. Specifically, the RDP setting with either the Kullback-Leibler (KL) divergence or Wasserstein-2 metric as the perception loss function is examined, and it is shown that for Gaussian vector sources, jointly Gaussian reconstructions are optimal. We further demonstrate that the optimal tradeoff can be expressed as an optimization problem, which can be explicitly solved. An interesting property of the optimal solution is as follows. Without the perception constraint, the traditional reverse water-filling solution for characterizing the rate-distortion (RD) tradeoff of a Gaussian vector source states that the optimal rate allocated to each component depends on a constant, called the water level. If the variance of a specific component is below the water level, it is assigned a zero compression rate. However, with active distortion and perception constraints, we show that the optimal rates allocated to the different components are always positive. Moreover, the water levels that determine the optimal rate allocation for different components are unequal. We further treat the special case of perceptually perfect reconstruction and study its RDP function in the high-distortion and low-distortion regimes to obtain insight to the structure of the optimal solution.

cs.IT↗

arXiv2022

On Distributed Lossy Coding of Symmetrically Correlated Gaussian Sources

A distributed lossy compression network with $L$ encoders and a decoder is considered. Each encoder observes a source and sends a compressed version to the decoder. The decoder produces a joint reconstruction of target signals with the mean squared error distortion below a given threshold. It is assumed that the observed sources can be expressed as the sum of target signals and corruptive noises which are independently generated from two symmetric multivariate Gaussian distributions. The minimum compression rate of this network versus the distortion threshold is referred to as the rate-distortion function, for which an explicit lower bound is established by solving a minimization problem. Our lower bound matches the well-known Berger-Tung upper bound for some values of the distortion threshold. The asymptotic gap between the upper and lower bounds is characterized in the large $L$ limit.

cs.IT↗

arXiv2021

Universal Rate-Distortion-Perception Representations for Lossy Compression

In the context of lossy compression, Blau & Michaeli (2019) adopt a mathematical notion of perceptual quality and define the information rate-distortion-perception function, generalizing the classical rate-distortion tradeoff. We consider the notion of universal representations in which one may fix an encoder and vary the decoder to achieve any point within a collection of distortion and perception constraints. We prove that the corresponding information-theoretic universal rate-distortion-perception function is operationally achievable in an approximate sense. Under MSE distortion, we show that the entire distortion-perception tradeoff of a Gaussian source can be achieved by a single encoder of the same rate asymptotically. We then characterize the achievable distortion-perception region for a fixed representation in the case of arbitrary distributions, identify conditions under which the aforementioned results continue to hold approximately, and study the case when the rate is not fixed in advance. This motivates the study of practical constructions that are approximately universal across the RDP tradeoff, thereby alleviating the need to design a new encoder for each objective. We provide experimental results on MNIST and SVHN suggesting that on image compression tasks, the operational tradeoffs achieved by machine learning models with a fixed encoder suffer only a small penalty when compared to their variable encoder counterparts.

cs.IT↗

arXiv2019

Stripe-based and Attribute-aware Network: A Two-Branch Deep Model for Vehicle Re-identification

Vehicle re-identification (Re-ID) has been attracting increasing interest in the field of computer vision due to the growing utilization of surveillance cameras in public security. However, vehicle Re-ID still suffers a similarity challenge despite the efforts made to solve this problem. This challenge involves distinguishing different instances with nearly identical appearances. In this paper, we propose a novel two-branch stripe-based and attribute-aware deep convolutional neural network (SAN) to learn the efficient feature embedding for vehicle Re-ID task. The two-branch neural network, consisting of stripe-based branch and attribute-aware branches, can adaptively extract the discriminative features from the visual appearance of vehicles. A horizontal average pooling and dimension-reduced convolutional layers are inserted into the stripe-based branch to achieve part-level features. Meanwhile, the attribute-aware branch extracts the global feature under the supervision of vehicle attribute labels to separate the similar vehicle identities with different attribute annotations. Finally, the part-level and global features are concatenated together to form the final descriptor of the input image for vehicle Re-ID. The final descriptor not only can separate vehicles with different attributes but also distinguish vehicle identities with the same attributes. The extensive experiments on both VehicleID and VeRi databases show that the proposed SAN method outperforms other state-of-the-art vehicle Re-ID approaches.

cs.CV↗