Video captioning evaluation aims at assessing the semantic consistency between video and candidate text, which should include measurement from two aspects: faithfulness (whether the information conveyed by candidate is correct w.r.t. video) and comprehensiveness (whether the main video content is covered by candidate). However, previous approaches have difficulty in evaluating faithfulness and comprehensiveness due to heavy reliance on references or heterogeneous of visual and textual data. In this paper, we propose a vision-involved evaluation metric based on a novel DuAl-Reconstruction Transformer, named DARTScore. DARTScore formulates the caption evaluation task as a dual-reconstruction problem to evaluate both faithfulness and comprehensiveness explicitly. Since the word in a candidate is usually related to several frames, DARTScore adaptively collects relevant frames to reconstruct the word and computes the reconstruction accuracy as faithfulness to inherently reflect whether the word information is contained in the video. In the inversive way, DARTScore reconstructs each frame with relevant words to evaluate comprehensiveness. By integrating fine-grained bidirectional reconstruction accuracies, DARTScore drills into each word in candidate and each frame in video to fully evaluate the semantic consistency. Furthermore, we collect and annotate two Chinese datasets with a large domain gap, named CRAETE-EVAL and VATEX-ZH-EVAL, to systematically evaluate existing metrics and fill the blank of Chinese video captioning evaluation. Experimental results show that DARTScore achieves higher correlation with human judgments, has lower reference reliance, and generalizes well to data from different domains.

Dominant pre-training works for image-text retrieval adopt “dual-encoder” architecture to enable high efficiency, where two encoders are used to extract image and text representations and contrastive learning is employed for global alignment. However, coarse-grained global alignment ignores detailed semantic associations between image and text. In this work, we propose a novel proxy task, named Visual-Language Error Modeling (ViLEM), to inject detailed image-text association into “dual-encoder” model by “proofreading” each word in the text against the corresponding image. Specifically, we first edit the image-paired text to automatically generate diverse plausible negative texts with pre-trained language models. ViLEM then enforces the model to discriminate the correctness of each word in the plausible negative texts and further correct the wrong words via resorting to image information. Further-more, we propose a multi-granularity interaction framework to perform ViLEM via interacting text features with both global and local image features, which associates local text semantics with both high-level visual context and multi-level local visual information. Our method surpasses state-of-the-art “dual-encoder” methods by a large margin on the image-text retrieval task and significantly improves discriminativeness to local textual semantics. Our model can also generalize well to video-text retrieval.

Mainstream Video-Language Pre-training (VLP) models [10, 26, 64] consist of three parts, a video encoder, a text encoder, and a video-text fusion Transformer. They pursue better performance via utilizing heavier unimodal encoders or multimodal fusion Transformers, resulting in increased parameters with lower efficiency in downstream tasks. In this work, we for the first time introduce an end-to-end VLP model, namely all-in-one Transformer, that embeds raw video and textual signals into joint representations using a unified backbone architecture. We argue that the unique temporal information of video data turns out to be a key barrier hindering the design of a modality-agnostic Transformer. To overcome the challenge, we introduce a novel and effective token rolling operation to encode temporal representations from video clips in a non-parametric manner. The careful design enables the representation learning of both video-text multimodal inputs and unimodal inputs using a unified model. Our pretrained ali-in-one Transformer is transferred to various downstream video-text tasks after fine-tuning, including text-video retrieval, video-question answering, multiple choice and video captioning. State-of-the-art performances with the minimal model FLOPs on ten datasets demonstrate the superiority of our method compared to the competitive counterparts. The code and pretrained models are available at https://github.com/showlab/all-in-one.

The state of the arts in vision-language pretraining (VLP) achieves exemplary performance but suffers from high training costs resulting from slow convergence and long training time, especially on large-scale web datasets. An essential obstacle to training efficiency lies in the entangled prediction rate (percentage of tokens for reconstruction) and corruption rate (percentage of corrupted tokens) in masked language modeling (MLM), that is, a proper corruption rate is achieved at the cost of a large portion of output tokens being excluded from prediction loss. To accelerate the convergence of VLP, we propose a new pretraining task, namely, free language modeling (FLM), that enables a 100% prediction rate with arbitrary corruption rates. FLM successfully frees the prediction rate from the tie-up with the corruption rate while allowing the corruption spans to be customized for each token to be predicted. FLM-trained models are encouraged to learn better and faster given the same GPU time by exploiting bidirectional contexts more flexibly. Extensive experiments show FLM could achieve an impressive 2.5 × pretraining time reduction in comparison to the MLM-based methods, while keeping competitive performance on both vision-language understanding and generation tasks. Code will be public at https://github.com/TencentARC/FLM.

Recent CLIP-guided 3D optimization methods, such as DreamFields [19] and PureCLIPNeRF [24], have achieved impressive results in zero-shot text-to-3D synthesis. However, due to scratch training and random initialization without prior knowledge, these methods often fail to generate accurate and faithful 3D structures that conform to the input text. In this paper, we make the first attempt to introduce explicit 3D shape priors into the CLIP-guided 3D optimization process. Specifically, we first generate a high-quality 3D shape from the input text in the text-to-shape stage as a 3D shape prior. We then use it as the initialization of a neural radiance field and optimize it with the full prompt. To address the challenging text-to-shape generation task, we present a simple yet effective approach that directly bridges the text and image modalities with a powerful text-to-image diffusion model. To narrow the style domain gap between the images synthesized by the text-to-image diffusion model and shape renderings used to train the image-to-shape generator, we further propose to jointly optimize a learnable text prompt and fine-tune the text-to-image diffusion model for rendering-style image generation. Our method, Dream3D, is capable of generating imaginative 3D content with superior visual quality and shape accuracy compared to state-of-the-art methods. Our project page is at https://bluestyle97.github.io/dream3d/.

Both masked image modeling (MIM) and natural language supervision have facilitated the progress of transferable visual pre-training. In this work, we seek the synergy between two paradigms and study the emerging properties when MIM meets natural language supervision. To this end, we present a novel masked visual Reconstruction In Language semantic Space (RILS) pre-training framework, in which sentence representations, encoded by the text encoder, serve as prototypes to transform the vision-only signals into patch-sentence probabilities as semantically meaningful MIM reconstruction targets. The vision models can therefore capture useful components with structured information by predicting proper semantic of masked tokens. Better visual representations could, in turn, improve the text encoder via the image-text alignment objective, which is essential for the effective MIM target transformation. Extensive experimental results demonstrate that our method not only enjoys the best of previous MIM and CLIP but also achieves further improvements on various tasks due to their mutual benefits. RILS exhibits advanced transferability on downstream classification, detection, and segmentation, especially for low-shot regimes. Code is available at https://github.com/hustvl/RILS.

Omnidirectional images (ODIs) have become increasingly popular, as their large field-of-view (FoV) can offer viewers the chance to freely choose the view directions in immersive environments such as virtual reality. The Möbius transformation is typically employed to further provide the opportunity for movement and zoom on ODIs, but applying it to the image level often results in blurry effect and aliasing problem. In this paper, we propose a novel deep learning-based approach, called OmniZoomer, to incorporate the Möbius transformation into the network for movement and zoom on ODIs. By learning various transformed feature maps under different conditions, the network is enhanced to handle the increasing edge curvatures, which alleviates the blurry effect. Moreover, to address the aliasing problem, we propose two key components. Firstly, to compensate for the lack of pixels for describing curves, we enhance the feature maps in the high-resolution (HR) space and calculate the transformed index map with a spatial index generation module. Secondly, considering that ODIs are inherently represented in the spherical space, we propose a spherical resampling module that combines the index map and HR feature maps to transform the feature maps for better spherical correlation. The transformed feature maps are decoded to output a zoomed ODI. Experiments show that our method can produce HR and high-quality ODIs with the flexibility to move and zoom in to the object of interest. Project page is available at http: //vlislab22.github.io/OmniZoomer/.

Despite the success in large-scale text-to-image generation and text-conditioned image editing, existing methods still struggle to produce consistent generation and editing results. For example, generation approaches usually fail to synthesize multiple images of the same objects/characters but with different views or poses. Meanwhile, existing editing methods either fail to achieve effective complex nonrigid editing while maintaining the overall textures and identity, or require time-consuming fine-tuning to capture the image-specific appearance. In this paper, we develop MasaCtrl, a tuning-free method to achieve consistent image generation and complex non-rigid image editing simultaneously. Specifically, MasaCtrl converts existing self-attention in diffusion models into mutual self-attention, so that it can query correlated local contents and textures from source images for consistency. To further alleviate the query confusion between foreground and background, we propose a mask-guided mutual self-attention strategy, where the mask can be easily extracted from the cross-attention maps. Extensive experiments show that the proposed MasaCtrl can produce impressive results in both consistent image generation and complex non-rigid real image editing.

We introduce HOSNeRF, a novel 360° free-viewpoint rendering method that reconstructs neural radiance fields for dynamic human-object-scene from a single monocular in-the-wild video. Our method enables pausing the video at any frame and rendering all scene details (dynamic humans, objects, and backgrounds) from arbitrary viewpoints. The first challenge in this task is the complex object motions in human-object interactions, which we tackle by introducing the new object bones into the conventional human skeleton hierarchy to effectively estimate large object deformations in our dynamic human-object model. The second challenge is that humans interact with different objects at different times, for which we introduce two new learnable object state embeddings that can be used as conditions for learning our human-object representation and scene representation, respectively. Extensive experiments show that HOSNeRF significantly outperforms SOTA approaches on two challenging datasets by a large margin of 40%~50% in terms of LPIPS. The code, data, and compelling examples of 360° free-viewpoint renderings from single videos: https://showlab.github.io/HOSNeRF.

Video Tagging intends to infer multiple tags spanning relevant content for a given video. Typically, video tags are freely defined and uploaded by a variety of users, so they have two characteristics: abundant in quantity and disordered intra-video. It is difficult for the existing multilabel classification and generation methods to adapt directly to this task. This paper proposes a novel generative model, Order-Prompted Tag Sequence Generation (OP-TSG), according to the above characteristics. It regards video tagging as a tag sequence generation problem guided by sample-dependent order prompts. These prompts are semantically aligned with tags and enable to decouple tag generation order, making the model focus on modeling the tag dependencies. Moreover, the word-based generation strategy enables the model to generate novel tags. To verify the effectiveness and generalization of the proposed method, a Chinese video tagging benchmark CREATE-tagging, and an English image tagging benchmark Pexel-tagging are established. Extensive results show that OP-TSG is significantly superior to other methods, especially the results on rare tags improve by 3.3% and 3% over SOTA methods on CREATE-tagging and Pexel-tagging, and novel tags generated on CREATE-tagging exhibit a tag gain of 7.04%.

To replicate the success of text-to-image (T2I) generation, recent works employ large-scale video datasets to train a text-to-video (T2V) generator. Despite their promising results, such paradigm is computationally expensive. In this work, we propose a new T2V generation setting—One-Shot Video Tuning, where only one text-video pair is presented. Our model is built on state-of-the-art T2I diffusion models pre-trained on massive image data. We make two key observations: 1) T2I models can generate still images that represent verb terms; 2) extending T2I models to generate multiple images concurrently exhibits surprisingly good content consistency. To further learn continuous motion, we introduce Tune-A-Video, which involves a tailored spatio-temporal attention mechanism and an efficient one-shot tuning strategy. At inference, we employ DDIM inversion to provide structure guidance for sampling. Extensive qualitative and numerical experiments demonstrate the remarkable ability of our method across various applications.

Cross-domain recommendation is an important method to improve recommender system performance, especially when observations in target domains are sparse. However, most existing techniques focus on single-target or dual-target cross-domain recommendation (CDR) and are hard to be generalized to CDR with multiple target domains. In addition, the negative transfer problem is prevalent in CDR, where the recommendation performance in a target domain may not always be enhanced by knowledge learned from a source domain, especially when the source domain has sparse data. In this study, we propose CAT-ART, a multi-target CDR method that learns to improve recommendations in all participating domains through representation learning and embedding transfer. Our method consists of two parts: a self-supervised Contrastive AuToencoder (CAT) framework to generate global user embeddings based on information from all participating domains, and an Attention-based Representation Transfer (ART) framework which transfers domain-specific user embeddings from other domains to assist with target domain recommendation. CAT-ART boosts the recommendation performance in any target domain through the combined use of the learned global user representation and knowledge transferred from other domains, in addition to the original user embedding in the target domain. We conducted extensive experiments on a collected real-world CDR dataset spanning 5 domains and involving a million users. Experimental results demonstrate the superiority of the proposed method over a range of prior arts. We further conducted ablation studies to verify the effectiveness of the proposed components. Our collected dataset will be open-sourced to facilitate future research in the field of multi-domain recommender systems and user modelling.

Few-shot action recognition has attracted increasing attention in recent years, but it remains challenging due to the intrinsic difficulty in learning transferable knowledge to generalize to novel classes by using a few labeled samples. Although some successful progress has been made, most few-shot action recognition methods commonly focus on the global characteristics of samples while ignoring the local characteristics of samples, which results in the weak generalization ability of the model. In this paper, we propose a task-aware dual-representation network (TADRNet) for few-shot action recognition, which learns how to adapt video representations to novel tasks in a meta-learning manner. It mainly includes a global relational graph subnetwork (GRG) and a fine-grained local representation subnetwork (FLR). Our method simultaneously considers both global and local characteristics of samples for few-shot action recognition. From a global perspective, we propose GRG to explore the relations across support-query sample pairs by using the relational graph neural network. To facilitate the few-shot visual learning, we propose a novel hybrid semantic attention module (HSA) for enhancing the discriminability of support and query features. From a local perspective, we utilize FLR to fully exploit the local characteristics of samples, which can improve the classification results obtained by GRG and thus guarantee high classification accuracy. Extensive experiments on four challenging benchmarks show that the proposed TADRNet significantly outperforms a variety of state-of-the-art few-shot action recognition methods.

Pretraining a model to learn transferable video-text representation for retrieval has attracted a lot of attention in recent years. Previous dominant works mainly adopt two separate encoders for efficient retrieval, but ignore local associations between videos and texts. Another line of research uses a joint encoder to interact video with texts, but results in low efficiency since each text-video pair needs to be fed into the model. In this work, we enable fine-grained video-text interactions while maintaining high efficiency for retrieval via a novel pretext task, dubbed as Multiple Choice Questions (MCQ), where a parametric module BridgeFormer is trained to answer the “questions” constructed by the text features via resorting to the video features. Specifically, we exploit the rich semantics of text (i.e., nouns and verbs) to build questions, with which the video encoder can be trained to capture more regional content and temporal dynamics. In the form of questions and answers, the semantic associations between local video-text features can be properly established. BridgeFormer is able to be removed for downstream retrieval, rendering an efficient and flexible model with only two encoders. Our method outperforms state-of-the-art methods on the popular text-to-video retrieval task in five datasets with different experimental setups (i.e., zero-shot andfine-tune), including HowTo100M (one million videos). We further conduct zero-shot action recognition, which can be cast as video-to-text retrieval, and our approach also significantly surpasses its counterparts. As an additional benefit, our method achieves competitive results with much shorter pre-training videos on single-modality downstream tasks, e.g., action recognition with linear evaluation.

Finding relevant moments and highlights in videos according to natural language queries is a natural and highly valuable common need in the current video content explosion era. Nevertheless, jointly conducting moment retrieval and highlight detection is an emerging research topic, even though its component problems and some related tasks have already been studied for a while. In this paper, we present the first unified framework, named Unified Multi-modal Transformers (UMT), capable of realizing such joint optimization while can also be easily degenerated for solving individual problems. As far as we are aware, this is the first scheme to integrate multi-modal (visual-audio) learning for either joint optimization or the individual moment retrieval task, and tackles moment retrieval as a keypoint detection problem using a novel query generator and query decoder. Extensive comparisons with existing methods and ablation studies on QVHighlights, Charades-STA, YouTube Highlights, and TVSum datasets demonstrate the effectiveness, superiority, and flexibility of the proposed method under various settings. Source code and pre-trained models are available at https://github.com/TencentARC/UMT.

Recently, by introducing large-scale dataset and strong transformer network, video-language pre-training has shown great success especially for retrieval. Yet, existing video-language transformer models do not explicitly fine-grained semantic align. In this work, we present Object-aware Transformers, an object-centric approach that extends video-language transformer to incorporate object representations. The key idea is to leverage the bounding boxes and object tags to guide the training process. We evaluate our model on three standard sub-tasks of video-text matching on four widely used benchmarks. We also provide deep analysis and detailed ablation about the proposed method. We show clear improvement in performance across all tasks and datasets considered, demonstrating the value of a model that incorporates object representations into a video-language architecture. The code has been released in https://github.com/FingerRec/OA-Transformer.

As a prerequisite of many text-related tasks such as text erasing and text style transfer, text segmentation arouses more and more attention recently. Current researches mainly focus on only English characters and digits, while few work studies Chinese characters due to the lack of pub-lic large-scale and high-quality Chinese datasets, which limits the practical application scenarios of text segmentation. Different from English which has a limited alphabet of letters, Chinese has much more basic characters with com-plex structures, making the problem more difficult to deal with. To better analyze this problem, we propose the Bi-lingual Text Segmentation (BTS) dataset, a benchmark that covers various common Chinese scenes including 14,250 diverse and fine-annotated text images. BTS mainly focuses on Chinese characters, and also contains English words and digits. We also introduce Prior Guided Text Segmen-tation Network (PGTSNet), the first baseline to handle bi-lingual and complex-structured text segmentation. A plug-in text region highlighting module and a text perceptual dis-criminator are proposed in PGTSNet to supervise the model with text prior, and guide for more stable and finer text seg-mentation. A variation loss is also employed for suppressing background noise under complex scene. Extensive ex-periments are conducted not only to demonstrate the neces-sity and superiority of the proposed dataset BTS, but also to show the effectiveness of the proposed PGTSNet compared with a variety of state-of-the-art text segmentation methods.

Modeling dynamic scenes is important for many applications such as virtual reality and telepresence. Despite achieving unprecedented fidelity for novel view synthesis in dynamic scenes, existing methods based on Neural Radiance Fields (NeRF) suffer from slow convergence (i.e., model training time measured in days). In this paper, we present DeVRF, a novel representation to accelerate learning dynamic radiance fields. The core of DeVRF is to model both the 3D canonical space and 4D deformation field of a dynamic, non-rigid scene with explicit and discrete voxel-based representations. However, it is quite challenging to train such a representation which has a large number of model parameters, often resulting in overfitting issues. To overcome this challenge, we devise a novel static-to-dynamic learning paradigm together with a new data capture setup that is convenient to deploy in practice. This paradigm unlocks efficient learning of deformable radiance fields via utilizing the 3D volumetric canonical space learnt from multi-view static images to ease the learning of 4D voxel deformation field with only few-view dynamic sequences. To further improve the efficiency of our DeVRF and its synthesized novel view's quality, we conduct thorough explorations and identify a set of strategies. We evaluate DeVRF on both synthetic and real-world dynamic scenes with different types of deformation. Experiments demonstrate that DeVRF achieves two orders of magnitude speedup (100× faster) with on-par high-fidelity results compared to the previous state-of-the-art approaches. The code and dataset are released in https://github.com/showlab/DeVRF.

Existing benchmark datasets for recommender systems (RS) either are created at a small scale or involve very limited forms of user feedback. RS models evaluated on such datasets often lack practical values for large-scale real-world applications. In this paper, we describe Tenrec, a novel and publicly available data collection for RS that records various user feedback from four different recommendation scenarios. To be specific, Tenrec has the following five characteristics: (1) it is large-scale, containing around 5 million users and 140 million interactions; (2) it has not only positive user feedback, but also true negative feedback (vs. one-class recommendation); (3) it contains overlapped users and items across four different scenarios; (4) it contains various types of user positive feedback, in forms of clicking, liking, sharing, and following, etc; (5) it contains additional features beyond the user IDs and item IDs. We verify Tenrec on ten diverse recommendation tasks by running several classical baseline models per task. Tenrec has the potential to become a useful benchmark dataset for a majority of popular recommendation tasks. Our source codes and datasets will be included in supplementary materials.

Users' behavioral predictions are crucially important for many domains including major e-commerce companies, ride-hailing platforms, social networking, and education. The success of such prediction strongly depends on the development of representation learning that can effectively model the dynamic evolution of user's behavior. This paper aims to develop a joint framework of combining inverse reinforcement learning (IRL) with deep learning (DL) regression model, called IRL-DL, to predict drivers' future behavior in ride-hailing platforms. Specifically, we formulate the dynamic evolution of each driver as a sequential decision-making problem and then employ IRL as representation learning to learn the preference vector of each driver. Then, we integrate drivers' preference vector with their static features (e.g., age, gender) and other attributes to build a regression model (e.g., LTSM-neural network) to predict drivers' future behavior. We use an extensive driver data set obtained from a ride-sharing platform to verify the effectiveness and efficiency of our IRL-DL framework, and results show that our IRL-DL framework can achieve consistent and remarkable improvements over models without drivers' preference vectors.

Nowadays, ridesharing has become one of the most popular services offered by online ride-hailing platforms (e.g., Uber and Didi Chuxing). Existing ridesharing platforms adopt the strategy that dispatches orders over the entire city at a uniform time interval. However, the uneven spatio-temporal order distributions in real-world ridesharing systems indicate that such an approach is suboptimal in practice. Thus, in this paper, we exploit adaptive dispatching intervals to boost the platform's profit under a guarantee of the maximum passenger waiting time. Specifically, we propose a hierarchical approach, which generates clusters of geographical areas suitable to share the same dispatching intervals, and then makes online decisions of selecting the appropriate time instances for order dispatch within each spatial cluster. Technically, we prove the impossibility of designing constant-competitive-ratio algorithms for the online adaptive interval problem, and propose online algorithms under partial or even zero future order knowledge that significantly improve the platform's profit over existing approaches. We conduct extensive experiments with a large-scale ridesharing order dataset, which contains all of the over 3.5 million ridesharing orders in Beijing, China, received by Didi Chuxing from October 1st to October 31st, 2018. The experimental results demonstrate that our proposed algorithms outperform existing approaches.

The estimated time of arrival (ETA) is a critical task in the intelligent transportation system, which involves the spatiotemporal data. Despite a significant amount of prior efforts have been made to design efficient and accurate systems for ETA task, few of them take structural graph data into account, much less the heterogeneous information network. In this paper, we propose HetETA to leverage heterogeneous information graph in ETA task. Specifically, we translate the road map into a multi-relational network and introduce a vehicle-trajectories based network to jointly consider the traffic behavior pattern. Moreover, we employ three components to model temporal information from recent periods, daily periods and weekly periods respectively. Each component comprises temporal convolutions and graph convolutions to learn representations of the spatiotemporal heterogeneous information for ETA task. Experiments on large-scale datasets illustrate the effectiveness of the proposed HetETA beyond the state-of-the-art methods, and show the importance of representation learning of heterogeneous information networks for ETA task.

Customer response prediction is critical in many industrial applications such as online advertising and recommendations. In particular, the challenge is greater for ride-hailing platforms such as Uber and DiDi, because the response prediction models need to consider historical and real-time event information in the physical environment, such as surrounding traffic and supply and demand conditions. In this paper, we propose to use dynamically constructed heterogeneous graph for each ongoing event to encode the attributes of the event and its surroundings. In addition, we propose a multi-layer graph neural network model to learn the impact of historical actions and the surrounding environment on the current events, and generate an effective event representation to improve the accuracy of the response model. We investigate this framework to two practical applications on the DiDi platform. Offline and online experiments show that the framework can significantly improve prediction performance. The framework has been deployed in the online production environment and serves tens of millions of event prediction requests every day.

E-hailing platforms have become an important component of public transportation in recent years. The supply (online drivers) and demand (passenger requests) are intrinsically imbalanced because of the pattern of human behavior, especially in time and locations such as peak hours and train stations. Hence, how to balance supply and demand is one of the key problems to satisfy passengers and drivers and increase social welfare. As an intuitive and effective approach to address this problem, driver repositioning has been employed by some real-world e-hailing platforms. In this paper, we describe a novel framework of driver repositioning system, which meets various requirements in practical situations, including robust driver experience satisfaction and multi-driver collaboration. We introduce an effective and user-friendly driver interaction design called “driver repositioning task”. A novel modularized algorithm is developed to generate the repositioning tasks in real time. To our knowledge, this is the first industry-level application of driver repositioning. We evaluate the proposed method in real-world experiments, achieving a 2% improvement of driver income. Our framework has been fully deployed in the online system of DiDi Chuxing and serves millions of drivers on a daily basis.

Point-of-interest (POI) retrieval that searches for relevant destination locations plays a significant role in on-demand ridehailing services. Existing solutions to POI retrieval mainly retrieve and rank POIs based on their semantic similarity scores. Although intuitive, quantifying the relevance of a Query-POI pair by single-field semantic similarity is subject to inherent limitations. In this paper, we propose a novel Query-POI relevance model for effective POI retrieval for ondemand ride-hailing services. Different from existing relevance models, we capture and represent multi-field and local&global semantic features of a Query-POI pair to measure the semantic similarity. Besides, we observe a hidden correlation between origin-destination locations in ride-hailing scenarios, and propose two location embeddings to characterize the specific correlation. By incorporating the geographic correlation with the semantic similarity, our model achieves better performance in POI ranking. Experimental results on two real-world click-through datasets demonstrate the improvements of our model over state-of-the-art methods.

This paper describes our effort to build extensible routers using a combination of general-purpose and network processors. We emphasize five overriding challenges that dictate our design decisions: (1) optimal resource allocation; (2) efficient but flexible scheduling of the CPU; (3) maintaining overall router robustness; (4) maximizing router performance; and (5) providing sufficient extensibility to enable the injection of new functionality into the router. We adopt a hierarchical architecture, in which packet flows traverse a range of processing/forwarding paths, thereby partitioning hardware and software in concert. This paper both presents the architecture, and describes our experiences implementing the architecture and addressing the five design challenges in a prototype built from Intel IXP 1200 and a Pentium. Copyright © 2005 John Wiley & Sons, Ltd.