QARM: Quantitative Alignment Multi-Modal Recommendation at Kuaishou

In recent years, with the significant evolution of multi-modal large models, many recommender researchers realized the potential of multi-modal information for user interest modeling. In industry, a wide-used modeling architecture is a cascading paradigm: (1) first pre-training a multi-modal model to provide omnipotent representations for downstream services; (2) The downstream recommendation model takes the multi-modal representation as additional input to fit real user-item behaviours. Although such paradigm achieves remarkable improvements, however, there still exist two problems that limit model performance: (1) Representation Unmatching: The pre-trained multi-modal model is always supervised by the classic NLP/CV tasks, while the recommendation models are supervised by real user-item interaction. As a result, the two fundamentally different tasks' goals were relatively separate, and there was a lack of consistent objective on their representations; (2) Representation Unlearning: The generated multi-modal representations are always stored in cache store and serve as extra fixed input of recommendation model, thus could not be updated by recommendation model gradient, further unfriendly for downstream training. Inspired by the two difficulties challenges in downstream tasks usage, we introduce a quantitative multi-modal framework to customize the specialized and trainable multi-modal information for different downstream models.

ChildMandarin: A Comprehensive Mandarin Speech Dataset for Young Children Aged 3-5

Automatic speech recognition (ASR) systems have advanced significantly with models like Whisper, Conformer, and self-supervised frameworks such as Wav2vec 2.0 and HuBERT. However, developing robust ASR models for young children's speech remains challenging due to differences in pronunciation, tone, and pace compared to adult speech. In this paper, we introduce a new Mandarin speech dataset focused on children aged 3 to 5, addressing the scarcity of resources in this area. The dataset comprises 41.25 hours of speech with carefully crafted manual transcriptions, collected from 397 speakers across various provinces in China, with balanced gender representation. We provide a comprehensive analysis of speaker demographics, speech duration distribution and geographic coverage. Additionally, we evaluate ASR performance on models trained from scratch, such as Conformer, as well as fine-tuned pre-trained models like HuBERT and Whisper, where fine-tuning demonstrates significant performance improvements. Furthermore, we assess speaker verification (SV) on our dataset, showing that, despite the challenges posed by the unique vocal characteristics of young children, the dataset effectively supports both ASR and SV tasks. This dataset is a valuable contribution to Mandarin child speech research and holds potential for applications in educational technology and child-computer interaction. It will be open-source and freely available for all academic purposes.

PB-LRDWWS System for the SLT 2024 Low-Resource Dysarthria Wake-Up Word Spotting Challenge

For the SLT 2024 Low-Resource Dysarthria Wake-Up Word Spotting (LRDWWS) Challenge, we introduce the PB-LRDWWS system. This system combines a dysarthric speech content feature extractor for prototype construction with a prototype-based classification method. The feature extractor is a fine-tuned HuBERT model obtained through a three-stage fine-tuning process using cross-entropy loss. This fine-tuned HuBERT extracts features from the target dysarthric speaker's enrollment speech to build prototypes. Classification is achieved by calculating the cosine similarity between the HuBERT features of the target dysarthric speaker's evaluation speech and prototypes. Despite its simplicity, our method demonstrates effectiveness through experimental results. Our system achieves second place in the final Test-B of the LRDWWS Challenge.

ELASTIC: Efficient Linear Attention for Sequential Interest Compression

State-of-the-art sequential recommendation models heavily rely on transformer's attention mechanism. However, the quadratic computational and memory complexities of self attention have limited its scalability for modeling users' long range behaviour sequences. To address this problem, we propose ELASTIC, an Efficient Linear Attention for SequenTial Interest Compression, requiring only linear time complexity and decoupling model capacity from computational cost. Specifically, ELASTIC introduces a fixed length interest experts with linear dispatcher attention mechanism which compresses the long-term behaviour sequences to a significantly more compact representation which reduces up to 90% GPU memory usage with x2.7 inference speed up. The proposed linear dispatcher attention mechanism significantly reduces the quadratic complexity and makes the model feasible for adequately modeling extremely long sequences. Moreover, in order to retain the capacity for modeling various user interests, ELASTIC initializes a vast learnable interest memory bank and sparsely retrieves compressed user's interests from the memory with a negligible computational overhead. The proposed interest memory retrieval technique significantly expands the cardinality of available interest space while keeping the same computational cost, thereby striking a trade-off between recommendation accuracy and efficiency. To validate the effectiveness of our proposed ELASTIC, we conduct extensive experiments on various public datasets and compare it with several strong sequential recommenders. Experimental results demonstrate that ELASTIC consistently outperforms baselines by a significant margin and also highlight the computational efficiency of ELASTIC when modeling long sequences. We will make our implementation code publicly available.

Moment&Cross: Next-Generation Real-Time Cross-Domain CTR Prediction for Live-Streaming Recommendation at Kuaishou

Kuaishou, is one of the largest short-video and live-streaming platform, compared with short-video recommendations, live-streaming recommendation is more complex because of: (1) temporarily-alive to distribution, (2) user may watch for a long time with feedback delay, (3) content is unpredictable and changes over time. Actually, even if a user is interested in the live-streaming author, it still may be an negative watching (e.g., short-view < 3s) since the real-time content is not attractive enough. Therefore, for live-streaming recommendation, there exists a challenging task: how do we recommend the live-streaming at right moment for users? Additionally, our platform's major exposure content is short short-video, and the amount of exposed short-video is 9x more than exposed live-streaming. Thus users will leave more behaviors on short-videos, which leads to a serious data imbalance problem making the live-streaming data could not fully reflect user interests. In such case, there raises another challenging task: how do we utilize users' short-video behaviors to make live-streaming recommendation better?

Enhancing Dysarthric Speech Recognition for Unseen Speakers via Prototype-Based Adaptation

Dysarthric speech recognition (DSR) presents a formidable challenge due to inherent inter-speaker variability, leading to severe performance degradation when applying DSR models to new dysarthric speakers. Traditional speaker adaptation methodologies typically involve fine-tuning models for each speaker, but this strategy is cost-prohibitive and inconvenient for disabled users, requiring substantial data collection. To address this issue, we introduce a prototype-based approach that markedly improves DSR performance for unseen dysarthric speakers without additional fine-tuning. Our method employs a feature extractor trained with HuBERT to produce per-word prototypes that encapsulate the characteristics of previously unseen speakers. These prototypes serve as the basis for classification. Additionally, we incorporate supervised contrastive learning to refine feature extraction. By enhancing representation quality, we further improve DSR performance, enabling effective personalized DSR. We release our code at https://github.com/NKU-HLT/PB-DSR.

MMBee: Live Streaming Gift-Sending Recommendations via Multi-Modal Fusion and Behaviour Expansion

Live streaming services are becoming increasingly popular due to real-time interactions and entertainment. Viewers can chat and send comments or virtual gifts to express their preferences for the streamers. Accurately modeling the gifting interaction not only enhances users' experience but also increases streamers' revenue. Previous studies on live streaming gifting prediction treat this task as a conventional recommendation problem, and model users' preferences using categorical data and observed historical behaviors. However, it is challenging to precisely describe the real-time content changes in live streaming using limited categorical information. Moreover, due to the sparsity of gifting behaviors, capturing the preferences and intentions of users is quite difficult. In this work, we propose MMBee based on real-time Multi-Modal Fusion and Behaviour Expansion to address these issues. Specifically, we first present a Multi-modal Fusion Module with Learnable Query (MFQ) to perceive the dynamic content of streaming segments and process complex multi-modal interactions, including images, text comments and speech. To alleviate the sparsity issue of gifting behaviors, we present a novel Graph-guided Interest Expansion (GIE) approach that learns both user and streamer representations on large-scale gifting graphs with multi-modal attributes. Comprehensive experiment results show that MMBee achieves significant performance improvements on both public datasets and Kuaishou real-world streaming datasets and the effectiveness has been further validated through online A/B experiments. MMBee has been deployed and is serving hundreds of millions of users at Kuaishou.

Physics-Aware Iterative Learning and Prediction of Saliency Map for Bimanual Grasp Planning

Learning the skill of human bimanual grasping can extend the capabilities of robotic systems when grasping large or heavy objects. However, it requires a much larger search space for grasp points than single-hand grasping and numerous bimanual grasping annotations for network learning, making both data-driven or analytical grasping methods inefficient and insufficient. We propose a framework for bimanual grasp saliency learning that aims to predict the contact points for bimanual grasping based on existing human single-handed grasping data. We learn saliency corresponding vectors through minimal bimanual contact annotations that establishes correspondences between grasp positions of both hands, capable of eliminating the need for training a large-scale bimanual grasp dataset. The existing single-handed grasp saliency value serves as the initial value for bimanual grasp saliency, and we learn a saliency adjusted score that adds the initial value to obtain the final bimanual grasp saliency value, capable of predicting preferred bimanual grasp positions from single-handed grasp saliency. We also introduce a physics-balance loss function and a physics-aware refinement module that enables physical grasp balance, capable of enhancing the generalization of unknown objects. Comprehensive experiments in simulation and comparisons on dexterous grippers have demonstrated that our method can achieve balanced bimanual grasping effectively.

ContentCTR: Frame-level Live Streaming Click-Through Rate Prediction with Multimodal Transformer

In recent years, live streaming platforms have gained immense popularity as they allow users to broadcast their videos and interact in real-time with hosts and peers. Due to the dynamic changes of live content, accurate recommendation models are crucial for enhancing user experience. However, most previous works treat the live as a whole item and explore the Click-through-Rate (CTR) prediction framework on item-level, neglecting that the dynamic changes that occur even within the same live room. In this paper, we proposed a ContentCTR model that leverages multimodal transformer for frame-level CTR prediction. First, we present an end-to-end framework that can make full use of multimodal information, including visual frames, audio, and comments, to identify the most attractive live frames. Second, to prevent the model from collapsing into a mediocre solution, a novel pairwise loss function with first-order difference constraints is proposed to utilize the contrastive information existing in the highlight and non-highlight frames. Additionally, we design a temporal text-video alignment module based on Dynamic Time Warping to eliminate noise caused by the ambiguity and non-sequential alignment of visual and textual information. We conduct extensive experiments on both real-world scenarios and public datasets, and our ContentCTR model outperforms traditional recommendation models in capturing real-time content changes. Moreover, we deploy the proposed method on our company platform, and the results of online A/B testing further validate its practical significance.

Self-Supervised Text Erasing with Controllable Image Synthesis

Recent efforts on scene text erasing have shown promising results. However, existing methods require rich yet costly label annotations to obtain robust models, which limits the use for practical applications. To this end, we study an unsupervised scenario by proposing a novel Self-supervised Text Erasing (STE) framework that jointly learns to synthesize training images with erasure ground-truth and accurately erase texts in the real world. We first design a style-aware image synthesis function to generate synthetic images with diverse styled texts based on two synthetic mechanisms. To bridge the text style gap between the synthetic and real-world data, a policy network is constructed to control the synthetic mechanisms by picking style parameters with the guidance of two specifically designed rewards. The synthetic training images with erasure ground-truth are then fed to train a coarse-to-fine erasing network. To produce better erasing outputs, a triplet erasure loss is designed to enforce the refinement stage to recover background textures. Moreover, we provide a new dataset (called PosterErase), which contains 60K high-resolution posters with texts and is more challenging for the text erasing task. The proposed method has been extensively evaluated with both PosterErase and the widely-used SCUT-Enstext dataset. Notably, on PosterErase, our unsupervised method achieves 5.07 in terms of FID, with a relative performance of 20.9% over existing supervised baselines.