Search

In the rapidly evolving field of e-commerce, the effectiveness of search re-ranking models is crucial for enhancing user experience and driving conversion rates. Despite significant advancements in feature representation and model architecture, the integration of multimodal information remains underexplored. This study addresses this gap by investigating the computation and fusion of textual and visual information in the context of re-ranking. We propose \textbf{A}dvancing \textbf{R}e-Ranking with \textbf{M}ulti\textbf{m}odal Fusion and \textbf{T}arget-Oriented Auxiliary Tasks (ARMMT), which integrates an attention-based multimodal fusion technique and an auxiliary ranking-aligned task to enhance item representation and improve targeting capabilities. This method not only enriches the understanding of product attributes but also enables more precise and personalized recommendations. Experimental evaluations on JD.com's search platform demonstrate that ARMMT achieves state-of-the-art performance in multimodal information integration, evidenced by a 0.22\% increase in the Conversion Rate (CVR), significantly contributing to Gross Merchandise Volume (GMV). This pioneering approach has the potential to revolutionize e-commerce re-ranking, leading to elevated user satisfaction and business growth.

We present Predictive Sparse Manifold Transform (PSMT), a minimalistic, interpretable and biologically plausible framework for learning and predicting natural dynamics. PSMT incorporates two layers where the first sparse coding layer represents the input sequence as sparse coefficients over an overcomplete dictionary and the second manifold learning layer learns a geometric embedding space that captures topological similarity and dynamic temporal linearity in sparse coefficients. We apply PSMT on a natural video dataset and evaluate the reconstruction performance with respect to contextual variability, the number of sparse coding basis functions and training samples. We then interpret the dynamic topological organization in the embedding space. We next utilize PSMT to predict future frames compared with two baseline methods with a static embedding space. We demonstrate that PSMT with a dynamic embedding space can achieve better prediction performance compared to static baselines. Our work establishes that PSMT is an efficient unsupervised generative framework for prediction of future visual stimuli., Comment: Paper presented at the 1st Workshop on High-dimensional Learning Dynamics (HLD) at the 40th International Conference on Machine Learning (ICML) 2023, Honolulu, Hawaii, USA (https://sites.google.com/view/hidimlearning), 10 pages

The violation of Bell inequality not only provides the most radical departure of quantum theory from classical concepts, but also paves the way of applications in such as device independent randomness certification. Here, we derive the tight upper bound of the maximum quantum value for chained Bell inequality with arbitrary number of measurements on each party. \lxh{ The constraints where the upper bound saturates are also presented. This method provides us the necessary and sufficient conditions for some quantum states to violate the chained Bell inequality with arbitrary number of measurements}. Based on the tight upper bound we present the lower bounds on the device independent randomness with respect to the Werner states. \lxh{In particular, we present lower bounds on the randomness generation rates of chained Bell inequality for different number of measurements, which are compared with the family of Bell inequalities proposed by Wooltorton et al. [Phys. Rev. Lett. 129, 150403 (2022)]. Our results show that chained Bell inequality with three measurements has certain advantages at a low level of noise and could be used to improve randomness generation rates in practice., Comment: 16 pages, 3 figures

Domain generalization in semantic segmentation aims to alleviate the performance degradation on unseen domains through learning domain-invariant features. Existing methods diversify images in the source domain by adding complex or even abnormal textures to reduce the sensitivity to domain specific features. However, these approaches depend heavily on the richness of the texture bank, and training them can be time-consuming. In contrast to importing textures arbitrarily or augmenting styles randomly, we focus on the single source domain itself to achieve generalization. In this paper, we present a novel adaptive texture filtering mechanism to suppress the influence of texture without using augmentation, thus eliminating the interference of domain-specific features. Further, we design a hierarchical guidance generalization network equipped with structure-guided enhancement modules, which purpose is to learn the domain-invariant generalized knowledge. Extensive experiments together with ablation studies on widely-used datasets are conducted to verify the effectiveness of the proposed model, and reveal its superiority over other state-of-the-art alternatives., Comment: Accepted by AAAI 2023

Self-testing, which refers to device independent characterization of the state and the measurement, enables the security of quantum information processing task certified independently of the operation performed inside the devices. Quantum states lie in the core of self-testing as key resources. However, for the different entangled states, usually different measurement settings should be taken in self-testing recipes. This may lead to the redundancy of measurement resources. In this work, we use fixed two-binary measurements and answer the question that what states can be self-tested with the same settings. By investigating the structure of generalized tilted-CHSH Bell operators with sum of squares decomposition method, we show that a family of two-qubit entangled states can be self-tested by the same measurement settings. The robustness analysis indicates that our scheme is feasible for practical experiment instrument. Moreover, our results can be applied to various quantum information processing tasks.

Deep learning has been widely applied in neuroimaging, including predicting brain-phenotype relationships from magnetic resonance imaging (MRI) volumes. MRI data usually requires extensive preprocessing prior to modeling, but variation introduced by different MRI preprocessing pipelines may lead to different scientific findings, even when using the identical data. Motivated by the data-centric perspective, we first evaluate how preprocessing pipeline selection can impact the downstream performance of a supervised learning model. We next propose two pipeline-invariant representation learning methodologies, MPSL and PXL, to improve robustness in classification performance and to capture similar neural network representations. Using 2000 human subjects from the UK Biobank dataset, we demonstrate that proposed models present unique and shared advantages, in particular that MPSL can be used to improve out-of-sample generalization to new pipelines, while PXL can be used to improve within-sample prediction performance. Both MPSL and PXL can learn more similar between-pipeline representations. These results suggest that our proposed models can be applied to mitigate pipeline-related biases, and to improve prediction robustness in brain-phenotype modeling., Comment: Extended Abstract presented at Machine Learning for Health (ML4H) symposium 2022, November 28th, 2022, New Orleans, United States & Virtual, http://www.ml4h.cc, 17 pages

This paper aims to introduce a robust singing voice synthesis (SVS) system to produce very natural and realistic singing voices efficiently by leveraging the adversarial training strategy. On one hand, we designed simple but generic random area conditional discriminators to help supervise the acoustic model, which can effectively avoid the over-smoothed spectrogram prediction and improve the expressiveness of SVS. On the other hand, we subtly combined the spectrogram with the frame-level linearly-interpolated F0 sequence as the input for the neural vocoder, which is then optimized with the help of multiple adversarial conditional discriminators in the waveform domain and multi-scale distance functions in the frequency domain. The experimental results and ablation studies concluded that, compared with our previous auto-regressive work, our new system can produce high-quality singing voices efficiently by fine-tuning different singing datasets covering from several minutes to a few hours. A large number of synthesized songs with different timbres are available online https://zzw922cn.github.io/wesinger2 and we highly recommend readers to listen to them., Comment: accepted at ICASSP 2023

In this paper, we develop a new multi-singer Chinese neural singing voice synthesis (SVS) system named WeSinger. To improve the accuracy and naturalness of synthesized singing voice, we design several specifical modules and techniques: 1) A deep bi-directional LSTM-based duration model with multi-scale rhythm loss and post-processing step; 2) A Transformer-alike acoustic model with progressive pitch-weighted decoder loss; 3) a 24 kHz pitch-aware LPCNet neural vocoder to produce high-quality singing waveforms; 4) A novel data augmentation method with multi-singer pre-training for stronger robustness and naturalness. To our knowledge, WeSinger is the first SVS system to adopt 24 kHz LPCNet and multi-singer pre-training simultaneously. Both quantitative and qualitative evaluation results demonstrate the effectiveness of WeSinger in terms of accuracy and naturalness, and WeSinger achieves state-of-the-art performance on the recent public Chinese singing corpus Opencpop\footnote{https://wenet.org.cn/opencpop/}. Some synthesized singing samples are available online\footnote{https://zzw922cn.github.io/wesinger/}., Comment: accepted at InterSpeech2022

This paper presents Nana-HDR, a new non-attentive non-autoregressive model with hybrid Transformer-based Dense-fuse encoder and RNN-based decoder for TTS. It mainly consists of three parts: Firstly, a novel Dense-fuse encoder with dense connections between basic Transformer blocks for coarse feature fusion and a multi-head attention layer for fine feature fusion. Secondly, a single-layer non-autoregressive RNN-based decoder. Thirdly, a duration predictor instead of an attention model that connects the above hybrid encoder and decoder. Experiments indicate that Nana-HDR gives full play to the advantages of each component, such as strong text encoding ability of Transformer-based encoder, stateful decoding without being bothered by exposure bias and local information preference, and stable alignment provided by duration predictor. Due to these advantages, Nana-HDR achieves competitive performance in naturalness and robustness on two Mandarin corpora.

In this work, a robust and efficient text-to-speech (TTS) synthesis system named Triple M is proposed for large-scale online application. The key components of Triple M are: 1) A sequence-to-sequence model adopts a novel multi-guidance attention to transfer complementary advantages from guiding attention mechanisms to the basic attention mechanism without in-domain performance loss and online service modification. Compared with single attention mechanism, multi-guidance attention not only brings better naturalness to long sentence synthesis, but also reduces the word error rate by 26.8%. 2) A new efficient multi-band multi-time vocoder framework, which reduces the computational complexity from 2.8 to 1.0 GFLOP and speeds up LPCNet by 2.75x on a single CPU.

Objectives: Precise segmentation of total extraocular muscles (EOM) and optic nerve (ON) is essential to assess anatomical development and progression of thyroid-associated ophthalmopathy (TAO). We aim to develop a semantic segmentation method based on deep learning to extract the total EOM and ON from orbital CT images in patients with suspected TAO. Materials and Methods: A total of 7,879 images obtained from 97 subjects who underwent orbit CT scans due to suspected TAO were enrolled in this study. Eighty-eight patients were randomly selected into the training/validation dataset, and the rest were put into the test dataset. Contours of the total EOM and ON in all the patients were manually delineated by experienced radiologists as the ground truth. A three-dimensional (3D) end-to-end fully convolutional neural network called semantic V-net (SV-net) was developed for our segmentation task. Intersection over Union (IoU) was measured to evaluate the accuracy of the segmentation results, and Pearson correlation analysis was used to evaluate the volumes measured from our segmentation results against those from the ground truth. Results: Our model in the test dataset achieved an overall IoU of 0.8207; the IoU was 0.7599 for the superior rectus muscle, 0.8183 for the lateral rectus muscle, 0.8481 for the medial rectus muscle, 0.8436 for the inferior rectus muscle and 0.8337 for the optic nerve. The volumes measured from our segmentation results agreed well with those from the ground truth (all R>0.98, P<0.0001). Conclusion: The qualitative and quantitative evaluations demonstrate excellent performance of our method in automatically extracting the total EOM and ON and measuring their volumes in orbital CT images. There is a great promise for clinical application to assess these anatomical structures for the diagnosis and prognosis of TAO., Comment: 17 pages, 8 figures

Entanglement lies at the heart of quantum mechanics, and has been identified an essential resource for diverse applications in quantum information. If entanglement could be verified without any trust in the devices of observers, i.e., in a device-independent (DI) way, then unconditional security can be guaranteed for various quantum information tasks. In this work, we propose an experimental-friendly DI protocol to certify the presence of entanglement, based on Einstein-Podolsky-Rosen (EPR) steering. We first establish the DI verification framework, relying on the measurement-device-independent technique and self-testing, and show it is able to verify all EPR-steerable states. In the context of three-measurement settings as per party, it is found to be noise robustness towards inefficient measurements and imperfect self-testing. Finally, a four-photon experiment is implemented to device-independently verify EPR-steering even for Bell local states. Our work paves the way for realistic implementations of secure quantum information tasks., Comment: Experiments are rerun to collect more data to do tomography and the text is significantly refined; Comments are still welcome

Self-testing refers to a device-independent way to uniquely identify the state and the measurement for uncharacterized quantum devices. The only information required comprises the number of measurements, the number of outputs of each measurement, and the statistics of each measurement. Earlier results on self-testing of multipartite state were restricted either to Dicke states or graph states. In this paper, we propose self-testing schemes for a large family of symmetric three-qubit states, namely the superposition of W state and GHZ state. We first propose and analytically prove a self-testing criterion for the special symmetric state with equal coefficients of the canonical basis, by designing subsystem self-testing of partially and maximally entangled state simultaneously. Then we demonstrate for the general case, the states can be self-tested numerically by the swap method combining semi-definite programming (SDP) in high precision., Comment: 9 pages,3 figures