Recent advances in vision-language models (VLMs) and reinforcement learning (RL) have driven progress in GUI automation. However, most existing methods rely on static, one-shot visual inputs and passive perception, lacking the ability to adaptively determine when, whether, and how to observe the interface. We present GUI-Eyes, a reinforcement learning framework for active visual perception in GUI tasks. To acquire more informative observations, the agent learns to make strategic decisions on both whether and how to invoke visual tools, such as cropping or zooming, within a two-stage reasoning process. To support this behavior, we introduce a progressive perception strategy that decomposes the decision-making into coarse exploration and fine-grained grounding, coordinated by a two-level policy. In addition, we design a spatially continuous reward function tailored to tool usage, which integrates both location proximity and region overlap to provide dense supervision and alleviate the reward sparsity common in GUI environments. On the ScreenSpot-Pro benchmark, GUI-Eyes-3B achieves 44.8% grounding accuracy using only 3k labeled samples, significantly outperforming both supervised and RL-based baselines. These results highlight that tool-aware active perception, enabled by staged policy reasoning and fine-grained reward feedback, is critical for building robust and data-efficient GUI agents.

With the rapid development of large multimodal models (LMMs), multimodal understanding applications are emerging. As most LMM inference requests originate from edge devices with limited computational capabilities, the predominant inference pipeline involves directly forwarding the input data to an edge server which handles all computations. However, this approach introduces high transmission latency due to limited uplink bandwidth of edge devices and significant computation latency caused by the prohibitive number of visual tokens, thus hindering delay-sensitive tasks and degrading user experience. To address this challenge, we propose a task-oriented feature compression (TOFC) method for multimodal understanding in a device-edge co-inference framework, where visual features are merged by clustering and encoded by a learnable and selective entropy model before feature projection. Specifically, we employ density peaks clustering based on$K$nearest neighbors to reduce the number of visual features, thereby minimizing both data transmission and computational complexity. Subsequently, a learnable entropy model with hyperprior is utilized to encode and decode merged features, further reducing transmission overhead. To enhance compression efficiency, multiple entropy models are adaptively selected based on the characteristics of the visual features, enabling a more accurate estimation of the probability distribution. Comprehensive experiments on seven visual question answering benchmarks validate the effectiveness of the proposed TOFC method. Results show that TOFC achieves up to 52% reduction in data transmission overhead and 63% reduction in system latency while maintaining identical task performance, compared with neural compression ELIC.

The recent development of 3D Gaussian splatting (3DGS) has led to great interest in 4D dynamic spatial reconstruction. Existing approaches mainly rely on full-length multi-view videos, while there has been limited exploration of online reconstruction methods that enable on-the-fly training and per-timestep streaming. Current 3DGS-based streaming methods treat the Gaussian primitives uniformly and constantly renew the densified Gaussians. Thus, they overlook the difference between dynamic and static features and neglect the temporal continuity of the scene. To address these limitations, we propose a novel pipeline for iterative streamable 4D dynamic spatial reconstruction. It comprises three stages: a selective inheritance stage that retains priors from previous timesteps to preserve the temporal continuity, a dynamics-aware shift stage that distinguishes dynamic and static primitives and employs distinct strategies to optimize their movements, and an error-guided densification stage that efficiently identifies Gaussians requiring densification to accommodate emerging objects. Our method achieves state-of-the-art performance in online 4D reconstruction, demonstrating compact storage, the fastest on-the-fly training speed, and superior representation quality.

Existing learning-based stereo image codec adopt sophisticated transformation with simple entropy models derived from single image codecs to encode latent representations. However, those entropy models struggle to effectively capture the spatial-disparity characteristics inherent in stereo images, which leads to suboptimal rate-distortion results. In this paper, we propose a stereo image compression framework, named CAMSIC. CAMSIC independently transforms each image to latent representation and employs a powerful decoder-free Transformer entropy model to capture both spatial and disparity dependencies, by introducing a novel content-aware masked image modeling (MIM) technique. Our content-aware MIM facilitates efficient bidirectional interaction between prior information and estimated tokens, which naturally obviates the need for an extra Transformer decoder. Experiments show that our stereo image codec achieves state-of-the-art rate-distortion performance on two stereo image datasets Cityscapes and InStereo2K with fast encoding and decoding speed.

Multimodal data is known to be helpful for visual tracking by improving robustness to appearance variations. However, sensor synchronization challenges often compromise data availability, particularly in video settings where shortages can be temporal. Despite its importance, this area remains underexplored. In this paper, we present the first comprehensive study on tracker performance with temporally incomplete multimodal data. Unsurprisingly, under such a circumstance, existing trackers exhibit significant performance degradation, as their rigid architectures lack the adaptability needed to effectively handle missing modalities. To address these limitations, we propose a flexible framework for robust multimodal tracking. We venture that a tracker should dynamically activate computational units based on missing data rates. This is achieved through a novel Heterogeneous Mixture-of-Experts fusion mechanism with adaptive complexity, coupled with a video-level masking strategy that ensures both temporal consistency and spatial completeness which is critical for effective video tracking. Surprisingly, our model not only adapts to varying missing rates but also adjusts to scene complexity. Extensive experiments show that our model achieves SOTA performance across 9 benchmarks, excelling in both conventional complete and missing modality settings. The code and benchmark will be publicly available at https://github.com/supertyd/FlexTrack/tree/main.

Background consistency remains a significant challenge in image editing tasks. Despite extensive developments, existing works still face a trade-off between maintaining similarity to the original image and generating content that aligns with the target. Here, we propose KV-Edit, a training-free approach that uses KV cache in DiTs to maintain background consistency, where background tokens are preserved rather than regenerated, eliminating the need for complex mechanisms or expensive training, ultimately generating new content that seamlessly integrates with the background within user-provided regions. We further explore the memory consumption of the KV cache during editing and optimize the space complexity to $O(1)$ using an inversion-free method. Our approach is compatible with any DiT-based generative model without additional training. Experiments demonstrate that KV-Edit significantly outperforms existing approaches in terms of both background and image quality, even surpassing training-based methods. Project webpage is available at https://xilluill.github.io/projectpages/KV-Edit

Fine-tuning Multimodal Large Language Models (MLLMs) in multi-task learning scenarios has emerged as an effective strategy for achieving cross-domain specialization. However, multi-task fine-tuning frequently induces performance degradation on open-response datasets. We posit that free-form answer generation primarily depends on language priors, and strengthening the integration of visual behavioral cues is critical for enhancing prediction robustness. In this work, we propose Noise Resilient Confidence Alignment to address the challenge of open-response overfitting during multi-task fine-tuning. Our approach prioritizes maintaining consistent prediction patterns in MLLMs across varying visual input qualities. To achieve this, we employ Gaussian perturbations to synthesize distorted visual inputs and enforce token prediction confidence alignment towards the normal visual branch. By explicitly linking confidence calibration to visual robustness, this method reduces over-reliance on language priors. We conduct extensive empirical evaluations across diverse multi-task downstream settings via popular MLLM architectures. The comprehensive experiment demonstrates the effectiveness of our method, showcasing its ability to alleviate open-response overfitting while maintaining satisfying multi-task fine-tuning performance.

Zijian Li, Qingyan Guo, Jiawei Shao, Lei Song, Jiang Bian, Jun Zhang, Rui Wang. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

Speculative decoding has become a widely adopted as an effective technique for lossless inference acceleration when deploying large language models (LLMs). While on-the-fly self-speculative methods offer seamless integration and broad utility, they often fall short of the speed gains achieved by methods relying on specialized training. Cascading a hierarchy of draft models promises further acceleration and flexibility, but the high cost of training multiple models has limited its practical application. In this paper, we propose a novel Cascade Adaptive Self-Speculative Decoding (CAS-Spec) method which constructs speculative draft models by leveraging dynamically switchable inference acceleration (DSIA) strategies, including layer sparsity and activation quantization. Furthermore, traditional vertical and horizontal cascade algorithms are inefficient when applied to self-speculative decoding methods. We introduce a Dynamic Tree Cascade (DyTC) algorithm that adaptively routes the multi-level draft models and assigns the draft lengths, based on the heuristics of acceptance rates and latency prediction. Our CAS-Spec method achieves state-of-the-art acceleration compared to existing on-the-fly speculative decoding methods, with an average speedup from $1.1\times$ to $2.3\times$ over autoregressive decoding across various LLMs and datasets. DyTC improves the average speedup by $47$\% and $48$\% over cascade-based baseline and tree-based baseline algorithms, respectively. CAS-Spec can be easily integrated into most existing LLMs and holds promising potential for further acceleration as self-speculative decoding techniques continue to evolve.

Large language models (LLMs) have demonstrated remarkable success across various domains. However, the substantial computational and memory demands pose significant challenges for deploying LLMs at the network edge. To address this issue, the existing studies mainly focus on either reducing the size of LLMs or distributing LLMs across multiple devices. However, the former approach suffers from performance degradation, while the latter incurs high communication cost due to the transmission of high-dimensional intermediate features. To tackle these issues, we propose a novel collaborative framework to support vision-language model (VLM) inference at the network edge, expanding speculative decoding into a multi-device scenario. In this framework, each edge device first utilizes its small VLM to generate draft tokens in an auto-regressive manner. These tokens are then transmitted to an edge server, where they are corrected in parallel by a large VLM. By benefiting from speculative decoding, the number of calls to the large VLM is reduced without degrading the inference performance. Furthermore, we minimize the average latency of inference tasks by developing a deep reinforcement learning algorithm to optimize the number of draft tokens generated at each iteration. Simulation results confirm that the proposed algorithm achieves a lower average latency compared to other baselines.

Task-oriented communication aims to extract and transmit task-relevant information to significantly reduce the communication overhead and transmission latency. However, theunpredictabledistribution shifts between training and test data, includingdomain shiftandsemantic shift, can dramatically undermine the system performance. In order to tackle these challenges, it is crucial to ensure that the encoded features can generalize todomain-shifteddata and detectsemantic-shifteddata, while remaining compact for transmission. In this paper, we propose a novel approach based on the information bottleneck (IB) principle and invariant risk minimization (IRM) framework. The proposed method aims to extract compact and informative features that possess high capability for effectivedomain-shift generalizationand accuratesemantic-shift detectionwithout any knowledge of the test data during training. Specifically, we propose an invariant feature encoding approach based on the IB principle and IRM framework fordomain-shiftgeneralization, which aims to find the causal relationship between the input data and task result by minimizing the complexity and domain dependence of the encoded feature. Furthermore, we enhance the task-oriented communication with the label-dependent feature encoding approach forsemantic-shift detectionwhich achieves joint gains in IB optimization and detection performance. To avoid the intractable computation of the IB-based objective, we leverage variational approximation to derive a tractable upper bound for optimization. Extensive simulation results on image classification tasks demonstrate that the proposed scheme outperforms state-of-the-art approaches and achieves a better rate-distortion tradeoff.

Federated learning (FL) is a distributed learning paradigm that maximizes the potential of data-driven models for edge devices without sharing their raw data. However, devices often have non-independent and identically distributed ( non-IID) data, meaning their local data distributions can vary significantly. The heterogeneity in input data distributions across devices, commonly referred to as the feature shift problem, can adversely impact the training convergence and accuracy of the global model. To analyze the intrinsic causes of the feature shift problem, we develop a generalization error bound in FL, which motivates us to propose FedCiR, a client-invariant representation learning framework that enables clients to extract informative and client-invariant features. Specifically, we improve the mutual information term between representations and labels to encourage representations to carry essential classification knowledge, and diminish the mutual information term between the client set and representations conditioned on labels to promote representations of clients to be client-invariant. We further incorporate two regularizers into the FL framework to bound the mutual information terms with an approximate global representation distribution to compensate for the absence of the ground-truth global representation distribution, thus achieving informative and client-invariant feature extraction. To achieve global representation distribution approximation, we propose a data-free mechanism performed by the server without compromising privacy. Extensive experiments demonstrate the effectiveness of our approach in achieving client-invariant representation learning and solving the data heterogeneity issue.

Federated learning (FL) has emerged as a privacy-preserving paradigm that trains neural networks on edge devices without collecting data at a central server. However, FL encounters an inherent challenge in dealing with non-independent and identically distributed (non-IID) data among devices. To address this challenge, this paper proposes a hard feature matching data synthesis (HFMDS) method to share auxiliary data besides local models. Specifically, synthetic data are generated by learning the essential class-relevant features of real samples and discarding the redundant features, which helps to effectively tackle the non-IID issue. For better privacy preservation, we propose a hard feature augmentation method to transfer real features towards the decision boundary, with which the synthetic data not only improve the model generalization but also erase the information of real features. By integrating the proposed HFMDS method with FL, we present a novel FL framework with data augmentation to relieve data heterogeneity. The theoretical analysis highlights the effectiveness of our proposed data synthesis method in solving the non-IID challenge. Simulation results further demonstrate that our proposed HFMDS-FL algorithm outperforms the baselines in terms of accuracy, privacy preservation, and complexity saving on various benchmark datasets.

With the development of artificial intelligence (AI) techniques and the increasing popularity of camera-equipped devices, many edge video analytics applications are emerging, calling for the deployment of computation-intensive AI models at the network edge. Edge inference is a promising solution to move computation-intensive workloads from low-end devices to a powerful edge server for video analytics, but device-server communications will remain a bottleneck due to limited bandwidth. This paper proposes a task-oriented communication framework for edge video analytics, where multiple devices collect the visual sensory data and transmit the informative features to an edge server for processing. To enable low-latency inference, this framework removes video redundancy in spatial and temporal domains and transmits minimal information that is essential for the downstream task, rather than reconstructing the videos on the edge server. Specifically, it extracts compact task-relevant features based on the deterministic information bottleneck (IB) principle, which characterizes a tradeoff between the informativeness of the features and the communication cost. As the features of consecutive frames are temporally correlated, we propose a temporal entropy model (TEM) to reduce the bitrate by taking the previous features as side information in feature encoding. To further improve the inference performance, we build a spatial-temporal fusion module on the server to integrate features of the current and previous frames for joint inference. Extensive experiments on video analytics tasks evidence that the proposed framework effectively encodes task-relevant information of video data and achieves a better rate-performance tradeoff than existing methods.

Federated learning (FL) has achieved great success as a privacy-preserving distributed training paradigm, where many edge devices collaboratively train a machine learning model by sharing the model updates instead of the raw data with a server. However, the heterogeneous computational and communication resources of edge devices give rise to stragglers that significantly decelerate the training process. To mitigate this issue, we propose a novel FL framework named stochastic coded federated learning (SCFL) that leverages coded computing techniques. In SCFL, before the training process starts, each edge device uploads a privacy-preserving coded dataset to the server, which is generated by adding Gaussian noise to the projected local dataset. During training, the server computes gradients on the global coded dataset to compensate for the missing model updates of the straggling devices. We design a gradient aggregation scheme to ensure that the aggregated model update is an unbiased estimate of the desired global update. Moreover, this aggregation scheme enables periodical model averaging to improve the training efficiency. We characterize the tradeoff between the convergence performance and privacy guarantee of SCFL. In particular, a more noisy coded dataset provides stronger privacy protection for edge devices but results in learning performance degradation. We further develop a contract-based incentive mechanism to coordinate such a conflict. The simulation results show that SCFL learns a better model within the given time and achieves a better privacy-performance tradeoff than the baseline methods. In addition, the proposed incentive mechanism grants better training performance than the conventional Stackelberg game approach.

Task-oriented communication is an emerging paradigm for next-generation communication networks, which extracts and transmits task-relevant information, instead of raw data, for downstream applications. Most existing deep learning (DL)-based task-oriented communication systems adopt a closed-world assumption, assuming either the same data distribution for training and testing, or the system could have access to a large out-of-distribution (OoD) dataset for retraining. However, in practical open-world scenarios, task-oriented communication systems will be exposed to unknown OoD data. The powerful approximation ability of learning methods may force the task-oriented communication systems to overfit the training data (i.e., in-distribution data). Therefore, these systems tend to provide overconfident judgments when encountering OoD data. Based on the information bottleneck (IB) framework, we propose a class conditional IB (CCIB) approach to address this problem, supported by information-theoretical insights. The idea is to extract distinguishable features from in-distribution data while keeping their compactness and informativeness. It is achieved by imposing the class conditional latent prior distribution and enforcing the latent of different classes to be far away from each other. Simulation results shall demonstrate that the proposed approach detects OoD data more efficiently than the baselines and state-of-the-art approaches, without compromising the rate-distortion tradeoff.

Prevalent predictive coding-based video compression methods rely on a heavy encoder to reduce temporal redundancy, which makes it challenging to deploy them on resource-constrained devices. Since the 1970s, distributed source coding theory has indicated that independent encoding and joint decoding with side information (SI) can achieve high-efficient compression of correlated sources. This has inspired a distributed coding architecture aiming at reducing the encoding complexity. However, traditional distributed coding methods suffer from a substantial performance gap to predictive coding ones. Inspired by the great success of learning-based compression, we propose the first end-to-end distributed deep video compression framework to improve the rate-distortion performance. A key ingredient is an effective SI generation module at the decoder, which helps to effectively exploit inter-frame correlations without computation-intensive encoder-side motion estimation and compensation. Experiments show that our method significantly outperforms conventional distributed video coding and H.264. Meanwhile, it enjoys 6 ∼ 7× encoding speedup against DVC [1] with comparable compression performance. Code is released at https://github.com/Xinjie-Q/Distributed-DVC.

Federated edge learning (FEEL) emerges as a privacy-preserving paradigm to effectively train deep learning models from the distributed data in 6G networks. Nevertheless, the limited coverage of a single edge server results in an insufficient number of participating client nodes, which may impair the learning performance. In this paper, we investigate a novel FEEL framework, namelysemi-decentralized federated edge learning(SD-FEEL), where multiple edge servers collectively coordinate a large number of client nodes. By exploiting the low-latency communication among edge servers for efficient model sharing, SD-FEEL incorporates more training data, while enjoying lower latency compared with conventional federated learning. We detail the training algorithm for SD-FEEL with three steps, including local model update, intra-cluster, and inter-cluster model aggregations. The convergence of this algorithm is proved on non-independent and identically distributed data, which reveals the effects of key parameters and provides design guidelines. Meanwhile, the heterogeneity of edge devices may cause the straggler effect and deteriorate the convergence speed of SD-FEEL. To resolve this issue, we propose an asynchronous training algorithm with a staleness-aware aggregation scheme, of which, the convergence is also analyzed. The simulations demonstrate the effectiveness and efficiency of the proposed algorithms for SD-FEEL and corroborate our analysis.

This paper investigates task-oriented communication for multi-device cooperative edge inference, where a group of distributed low-end edge devices transmit the extracted features of local samples to a powerful edge server for inference. While cooperative edge inference can overcome the limited sensing capability of a single device, it substantially increases the communication overhead and may incur excessive latency. To enable low-latency cooperative inference, we propose a learning-based communication scheme that optimizes local feature extraction and distributed feature encoding in a task-oriented manner, i.e., to remove data redundancy and transmit information that is essential for the downstream inference task rather than reconstructing the data samples at the edge server. Specifically, we leverage Tishby’s information bottleneck (IB) principle (Tishby et al., 2000) to extract the task-relevant feature at each edge device, and adopt the distributed information bottleneck (DIB) framework of Aguerri and Zaidi, 2021, to formalize a single-letter characterization of the optimal rate-relevance tradeoff for distributed feature encoding. To admit flexible control of the communication overhead, we extend the DIB framework to a distributed deterministic information bottleneck (DDIB) objective that explicitly incorporates the representational costs of the encoded features. As the IB-based objectives are computationally prohibitive for high-dimensional data, we adopt variational approximations to make the optimization problems tractable. To compensate for the potential performance loss due to the variational approximations, we also develop a selective retransmission (SR) mechanism to identify the redundancy in the encoded features among multiple edge devices to attain additional communication overhead reduction. Extensive experiments on multi-view image classification and multi-view object recognition tasks evidence that the proposed task-oriented communication scheme achieves a better rate-relevance tradeoff than existing methods.

Federated edge learning (FEEL) has drawn much attention as a privacy-preserving distributed learning framework for mobile edge networks. In this work, we investigate a novel semi-decentralized FEEL (SD-FEEL) architecture where multiple edge servers collaborate to incorporate more data from edge devices in training. Despite the low training latency enabled by fast edge aggregation, the device heterogeneity in computational resources deteriorates the efficiency. This paper proposes an asynchronous training algorithm to overcome this issue in SD-FEEL, where edge servers are allowed to independently set deadlines for the associated client nodes and trigger the model aggregation. To deal with different levels of model staleness, we design a staleness-aware aggregation scheme and analyze its convergence. Simulation results demonstrate the effectiveness of our proposed algorithm in achieving faster convergence and better learning performance than synchronous training.

Federated learning (FL) has attracted much attention as a privacy-preserving distributed machine learning framework, where many clients collaboratively train a machine learning model by exchanging model updates with a parameter server instead of sharing their raw data. Nevertheless, FL training suffers from slow convergence and unstable performance due to stragglers caused by the heterogeneous computational resources of clients and fluctuating communication rates. This paper proposes a coded FL framework to mitigate the straggler issue, namely stochastic coded federated learning (SCFL). In this framework, each client generates a privacy-preserving coded dataset by adding additive noise to the random linear combination of its local data. The server collects the coded datasets from all the clients to construct a composite dataset, which helps to compensate for the straggling effect. In the training process, the server as well as clients perform mini-batch stochastic gradient descent (SGD), and the server adds a make-up term in model aggregation to obtain unbiased gradient estimates. We characterize the privacy guarantee by the mutual information differential privacy (MI-DP) and analyze the convergence performance in federated learning. Besides, we demonstrate a privacy-performance tradeoff of the proposed SCFL method by analyzing the influence of the privacy constraint on the convergence rate. Finally, numerical experiments corroborate our analysis and show the benefits of SCFL in achieving fast convergence while preserving data privacy.

Federated learning (FL) strives to enable collaborative training of machine learning models without centrally collecting clients' private data. Different from centralized training, the local datasets across clients in FL are non-independent and identically distributed (non-IID). In addition, the data-owning clients may drop out of the training process arbitrarily. These characteristics will significantly degrade the training performance. This paper proposes a Dropout-Resilient Secure Federated Learning (DReS-FL) framework based on Lagrange coded computing (LCC) to tackle both the non-IID and dropout problems. The key idea is to utilize Lagrange coding to secretly share the private datasets among clients so that each client receives an encoded version of the global dataset, and the local gradient computation over this dataset is unbiased. To correctly decode the gradient at the server, the gradient function has to be a polynomial in a finite field, and thus we construct polynomial integer neural networks (PINNs) to enable our framework. Theoretical analysis shows that DReS-FL is resilient to client dropouts and provides privacy protection for the local datasets. Furthermore, we experimentally demonstrate that DReS-FL consistently leads to significant performance gains over baseline methods.

Federated edge learning (FEEL) has emerged as an effective approach to reduce the large communication latency in Cloud-based machine learning solutions, while preserving data privacy. Unfortunately, the learning performance of FEEL may be compromised due to limited training data in a single edge cluster. In this paper, we investigate a novel framework of FEEL, namely semi-decentralized federated edge learning (SD-FEEL). By allowing model aggregation across different edge clusters, SD-FEEL enjoys the benefit of FEEL in reducing the training latency, while improving the learning performance by accessing richer training data from multiple edge clusters. A training algorithm for SD-FEEL with three main procedures in each round is presented, including local model updates, intra-cluster and inter-cluster model aggregations, which is proved to converge on non-independent and identically distributed (non-IID) data. We also characterize the interplay between the network topology of the edge servers and the communication overhead of inter-cluster model aggregation on the training performance. Experiment results corroborate our analysis and demonstrate the effectiveness of SD-FFEL in achieving faster convergence than traditional federated learning architectures. Besides, guidelines on choosing critical hyper-parameters of the training algorithm are also provided.

This paper investigates task-oriented communication for edge inference, where a low-end edge device transmits the extracted feature vector of a local data sample to a powerful edge server for processing. It is critical to encode the data into aninformativeandcompactrepresentation for low-latency inference given the limited bandwidth. We propose a learning-based communication scheme that jointly optimizes feature extraction, source coding, and channel coding in a task-oriented manner, i.e., targeting the downstream inference task rather than data reconstruction. Specifically, we leverage an information bottleneck (IB) framework to formalize a rate-distortion tradeoff between the informativeness of the encoded feature and the inference performance. As the IB optimization is computationally prohibitive for the high-dimensional data, we adopt a variational approximation, namely the variational information bottleneck (VIB), to build a tractable upper bound. To reduce the communication overhead, we leverage a sparsity-inducing distribution as the variational prior for the VIB framework to sparsify the encoded feature vector. Furthermore, considering dynamic channel conditions in practical communication systems, we propose a variable-length feature encoding scheme based on dynamic neural networks to adaptively adjust the activated dimensions of the encoded feature to different channel conditions. Extensive experiments evidence that the proposed task-oriented communication system achieves a better rate-distortion tradeoff than baseline methods and significantly reduces the feature transmission latency in dynamic channel conditions.

Device-edge co-inference, which partitions a deep neural network between a resource-constrained mobile device and an edge server, recently emerges as a promising paradigm to support intelligent mobile applications. To accelerate the in-ference process, on-device model sparsification and intermediate feature compression are regarded as two prominent techniques. However, as the on-device model sparsity level and intermediate feature compression ratio have direct impacts on computation workload and communication overhead respectively, and both of them affect the inference accuracy, finding the optimal values of these hyper-parameters brings a major challenge due to the large search space. In this paper, we endeavor to develop an efficient algorithm to determine these hyper-parameters. By selecting a suitable model split point and a pair of encoder/decoder for the intermediate feature vector, this problem is casted as a sequential decision problem, for which, a novel automated machine learning (AutoML) framework is proposed based on deep reinforcement learning (DRL). Experiment results on an image classification task demonstrate the effectiveness of the proposed framework in achieving a better communication-computation trade-off and significant inference speedup against various baseline schemes.

The recent advancements of three-dimensional (3D) data acquisition devices have spurred a new breed of applications that rely on point cloud data processing. However, processing a large volume of point cloud data brings a significant workload on resource-constrained mobile devices, prohibiting from unleashing their full potentials. Built upon the emerging paradigm of device-edge co-inference, where an edge device extracts and transmits the intermediate feature to an edge server for further processing, we propose Branchy-GNN for efficient graph neural network (GNN) based point cloud processing by leveraging edge computing platforms. In order to reduce the on-device computational cost, the Branchy-GNN adds branch networks for early exiting. Besides, it employs learning-based joint source-channel coding (JSCC) for the intermediate feature compression to reduce the communication overhead. Our experimental results demonstrate that the proposed Branchy-GNN secures a significant latency reduction compared with several benchmark methods.