Shared Embedding Space Research Articles

A deep learning method to integrate extracelluar miRNA with mRNA for cancer studies.

Extracellular miRNAs (exmiRs) and intracellular mRNAs both can serve as promising biomarkers and therapeutic targets for various diseases. However, exmiR expression data is often noisy, and obtaining intracellular mRNA expression data usually involves intrusive procedures. To gain valuable insights into disease mechanisms, it is thus essential to improve the quality of exmiR expression data and develop noninvasive methods for assessing intracellular mRNA expression. We developed CrossPred, a deep-learning multi-encoder model for the cross-prediction of exmiRs and mRNAs. Utilizing contrastive learning, we created a shared embedding space to integrate exmiRs and mRNAs. This shared embedding was then used to predict intracellular mRNA expression from noisy exmiR data and to predict exmiR expression from intracellular mRNA data. We evaluated CrossPred on three types of cancers and assessed its effectiveness in predicting the expression levels of exmiRs and mRNAs. CrossPred outperformed the baseline encoder-decoder model, exmiR or mRNA-based models, and variational autoencoder models. Moreover, the integration of exmiR and mRNA data uncovered important exmiRs and mRNAs associated with cancer. Our study offers new insights into the bidirectional relationship between mRNAs and exmiRs. The datasets and tool are available at https://doi.org/10.5281/zenodo.13891508.

BCAN: Bidirectional Correct Attention Network for Cross-Modal Retrieval.

As a fundamental topic in bridging the gap between vision and language, cross-modal retrieval purposes to obtain the correspondences' relationship between fragments, i.e., subregions in images and words in texts. Compared with earlier methods that focus on learning the visual semantic embedding from images and sentences to the shared embedding space, the existing methods tend to learn the correspondences between words and regions via cross-modal attention. However, such attention-based approaches invariably result in semantic misalignment between subfragments for two reasons: 1) without modeling the relationship between subfragments and the semantics of the entire images or sentences, it will be hard for such approaches to distinguish images or sentences with multiple same semantic fragments and 2) such approaches focus attention evenly on all subfragments, including nonvisual words and a lot of redundant regions, which also will face the problem of semantic misalignment. To solve these problems, this article proposes a bidirectional correct attention network (BCAN), which introduces a novel concept of the relevance between subfragments and the semantics of the entire images or sentences and designs a novel correct attention mechanism by modeling the local and global similarity between images and sentences to correct the attention weights focused on the wrong fragments. Specifically, we introduce a concept about the semantic relationship between subfragments and entire images or sentences and use this concept to solve the semantic misalignment from two aspects. In our correct attention mechanism, we design two independent units to correct the weight of attention focused on the wrong fragments. Global correct unit (GCU) with modeling the global similarity between images and sentences into the attention mechanism to solve the semantic misalignment problem caused by focusing attention on relevant subfragments in irrelevant pairs (RI) and the local correct unit (LCU) consider the difference in the attention weights between fragments among two steps to solve the semantic misalignment problem caused by focusing attention on irrelevant subfragments in relevant pairs (IR). Extensive experiments on large-scale MS-COCO and Flickr30K show that our proposed method outperforms all the attention-based methods and is competitive to the state-of-the-art. Our code and pretrained model are publicly available at: https://github.com/liuyyy111/BCAN.

Boosting cross-modal retrieval in remote sensing via a novel unified attention network

With the rapid advent and abundance of remote sensing data in different modalities, cross-modal retrieval tasks have gained importance in the research community. Cross-modal retrieval belongs to the research paradigm in which the query is of one modality and the retrieved output is of the other modality. In this paper, the remote sensing (RS) data modalities considered are the earth observation optical data (aerial photos) and the corresponding hand-drawn sketches. The main challenge of the cross-modal retrieval research objective for optical remote sensing images and the corresponding sketches is the distribution gap between the shared embedding space of the modalities. Prior attempts to resolve this issue have not yielded satisfactory outcomes regarding accurately retrieving cross-modal sketch-image RS data. The state-of-the-art architectures used conventional convolutional architectures, which focused on local pixel-wise information about the modalities to be retrieved. This limits the interaction between the sketch texture and the corresponding image, making these models susceptible to overfitting datasets with particular scenarios. To circumvent this limitation, we suggest establishing multi-modal correspondence using a novel architecture of the combined self and cross-attention algorithms, SPCA-Net to minimize the modality gap by employing attention mechanisms for the query and other modalities. Efficient cross-modal retrieval is achieved through the suggested attention architecture, which empirically emphasizes the global information of the relevant query modality and bridges the domain gap through a unique pairwise cross-attention network. In addition to the novel architecture, this paper introduces a unique loss function, label-specific supervised contrastive loss, tailored to the intricacies of the task and to enhance the discriminative power of the learned embeddings. Extensive evaluations are conducted on two sketch-image remote sensing datasets, Earth-on-Canvas and RSketch. Under the same experimental conditions, the performance metrics of our proposed model beat the state-of-the-art architectures by significant margins of 16.7%, 18.9%, 33.7%, and 40.9% correspondingly.

Subject-independent auditory spatial attention detection based on brain topology modeling and feature distribution alignment

Auditory spatial attention detection (ASAD) seeks to determine which speaker in a surround sound field a listener is focusing on based on the one’s brain biosignals. Although existing studies have achieved ASAD from a single-trial electroencephalogram (EEG), the huge inter-subject variability makes them generally perform poorly in cross-subject scenarios. Besides, most ASAD methods do not take full advantage of topological relationships between EEG channels, which are crucial for high-quality ASAD. Recently, some advanced studies have introduced graph-based brain topology modeling into ASAD, but how to calculate edge weights in a graph to better capture actual brain connectivity is worthy of further investigation. To address these issues, we propose a new ASAD method in this paper. First, we model a multi-channel EEG segment as a graph, where differential entropy serves as the node feature, and a static adjacency matrix is generated based on inter-channel mutual information to quantify brain functional connectivity. Then, different subjects’ EEG graphs are encoded into a shared embedding space through a total variation graph neural network. Meanwhile, feature distribution alignment based on multi-kernel maximum mean discrepancy is adopted to learn subject-invariant patterns. Note that we align EEG embeddings of different subjects to reference distributions rather than align them to each other for the purpose of privacy preservation. A series of experiments on open datasets demonstrate that the proposed model outperforms state-of-the-art ASAD models in cross-subject scenarios with relatively low computational complexity, and feature distribution alignment improves the generalizability of the proposed model to a new subject.

A shared model-based linguistic space for transmitting our thoughts from brain to brain in natural conversations

Effective communication hinges on a mutual understanding of word meaning in different contexts. We recorded brain activity using electrocorticography during spontaneous, face-to-face conversations in five pairs of epilepsy patients. We developed a model-based coupling framework that aligns brain activity in both speaker and listener to a shared embedding space from a large language model (LLM). The context-sensitive LLM embeddings allow us to track the exchange of linguistic information, word by word, from one brain to another in natural conversations. Linguistic content emerges in the speaker's brain before word articulation and rapidly re-emerges in the listener's brain after word articulation. The contextual embeddings better capture word-by-word neural alignment between speaker and listener than syntactic and articulatory models. Our findings indicate that the contextual embeddings learned by LLMs can serve as an explicit numerical model of the shared, context-rich meaning space humans use to communicate their thoughts to one another.

Art and Language After AI

By ingesting a vast corpus of source material, generative deep learning models are capable of encoding multi-modal data into a shared embedding space, producing synthetic outputs which cannot be decomposed into their constituent parts. These models call into question the relation of conceptualisation and production in creative practices spanning musical composition to visual art. Moreover, artificial intelligence as a research program poses deeper questions regarding the very nature of aesthetic categories and their constitution. In this essay I will consider the intelligibility of the art object through the lens of a particular family of machine learning models, known as ‘latent diffusion’, extending an aesthetic theory to complement the image of thought the models (re)present to us. This will lead to a discussion on the semantics of computational states, probing the inferential and referential capacities of said models. Throughout I will endorse a topological view of computation, which will inform the neural turn in computer science, characterised as a shift from the notion of a stored program to that of a cognitive model. Lastly, I will look at the instability of these models by analysing their limitations in terms of compositionality and grounding.

Source-free domain adaptation for semantic image segmentation using internal representations.

Semantic segmentation models trained on annotated data fail to generalize well when the input data distribution changes over extended time period, leading to requiring re-training to maintain performance. Classic unsupervised domain adaptation (UDA) attempts to address a similar problem when there is target domain with no annotated data points through transferring knowledge from a source domain with annotated data. We develop an online UDA algorithm for semantic segmentation of images that improves model generalization on unannotated domains in scenarios where source data access is restricted during adaptation. We perform model adaptation by minimizing the distributional distance between the source latent features and the target features in a shared embedding space. Our solution promotes a shared domain-agnostic latent feature space between the two domains, which allows for classifier generalization on the target dataset. To alleviate the need of access to source samples during adaptation, we approximate the source latent feature distribution via an appropriate surrogate distribution, in this case a Gaussian mixture model (GMM).

Federated learning with knowledge distillation for multi-organ segmentation with partially labeled datasets

The state-of-the-art multi-organ CT segmentation relies on deep learning models, which only generalize when trained on large samples of carefully curated data. However, it is challenging to train a single model that can segment all organs and types of tumors since most large datasets are partially labeled or are acquired across multiple institutes that may differ in their acquisitions. A possible solution is Federated learning, which is often used to train models on multi-institutional datasets where the data is not shared across sites. However, predictions of federated learning can be unreliable after the model is locally updated at sites due to ‘catastrophic forgetting’. Here, we address this issue by using knowledge distillation (KD) so that the local training is regularized with the knowledge of a global model and pre-trained organ-specific segmentation models. We implement the models in a multi-head U-Net architecture that learns a shared embedding space for different organ segmentation, thereby obtaining multi-organ predictions without repeated processes. We evaluate the proposed method using 8 publicly available abdominal CT datasets of 7 different organs. Of those datasets, 889 CTs were used for training, 233 for internal testing, and 30 volumes for external testing. Experimental results verified that our proposed method substantially outperforms other state-of-the-art methods in terms of accuracy, inference time, and the number of parameters.

Zero-Shot Learning With Attentive Region Embedding and Enhanced Semantics.

The performance of zero-shot learning (ZSL) can be improved progressively by learning better features and generating pseudosamples for unseen classes. Existing ZSL works typically learn feature extractors and generators independently, which may shift the unseen samples away from their real distribution and suffers from the domain bias problem. In this article, to tackle this challenge, we propose a variational autoencoder (VAE)-based framework, that is, joint Attentive Region Embedding with Enhanced Semantics (AREES), which is tailored to advance the zero-shot recognition. Specifically, AREES is end-to-end trainable and consists of three network branches: 1) attentive region embedding is used to learn the semantic-guided visual features by the attention mechanism (AM); 2) a decomposition structure and a semantic pivot regularization are used to extract enhanced semantics; and 3) a multimodal VAE (mVAE) with the cross-reconstruction loss and the distribution alignment loss is used to obtain a shared latent embedding space of visual features and semantics. Finally, features' extraction and features' generation are optimized together in AREES to address the domain shift problem to a large extent. The comprehensive evaluations on six benchmarks, including the ImageNet, demonstrate the superiority of the proposed model over its state-of-the-art counterparts.

Visual-audio correspondence and its effect on video tipping: Evidence from Bilibili vlogs

Video tipping takes a remarkable share in the income of online streaming platforms such as Bilibili. There are some specific mappings between the audio and visual signals that viewers can sense (e.g., congruency of pitch and size), which is generally called visual-audio correspondence (VAC). And it is believed to influence viewer satisfaction with video clips. The way to automatically measure VAC, however, still remains missing and its possible effect on video tipping is rarely examined in previous efforts. In this study, a deep neural network with two sub-networks, namely VAC-Net, is established to map both visual and audio stimuli into a shared embedding space. And the Euclidean distance between visual and audio representations in this space is accordingly presented to be the indicator of VAC. Pre-trained models of both modalities and the triplet loss are further leveraged to train the VAC-Net and it competently evaluates VAC of video clips with a test accuracy of 68.37% by outperforming alternative baselines and even exceeding humans on the similar task. Lab-experiments further show that the VAC measurement of VAC-Net conforms to human cognition. Second, considering that viewers’ tipping behavior (TIP) on videos is consistent with the pricing strategy Pay What You Want (PWYW), it is hypothesized that VAC would indirectly influence TIP by reshaping viewer satisfaction (VS). Regression models are thus built to test the hypotheses and it is found that VAC can promote TIP by enhancing VS significantly. Additional tests also demonstrate the robustness of this mechanism by considering various controls and measurement errors. Our results supplement PWYW in streaming videos with a new motive of VAC for viewer tipping and provide streaming practitioners with an automatic tool to estimate the tips videos will receive.

A multi-band centroid contrastive reconstruction fusion network for motor imagery electroencephalogram signal decoding.

Motor imagery (MI) brain-computer interface (BCI) assist users in establishing direct communication between their brain and external devices by decoding the movement intention of human electroencephalogram (EEG) signals. However, cerebral cortical potentials are highly rhythmic and sub-band features, different experimental situations and subjects have different categories of semantic information in specific sample target spaces. Feature fusion can lead to more discriminative features, but simple fusion of features from different embedding spaces leading to the model global loss is not easily convergent and ignores the complementarity of features. Considering the similarity and category contribution of different sub-band features, we propose a multi-band centroid contrastive reconstruction fusion network (MB-CCRF). We obtain multi-band spatio-temporal features by frequency division, preserving the task-related rhythmic features of different EEG signals; use a multi-stream cross-layer connected convolutional network to perform a deep feature representation for each sub-band separately; propose a centroid contrastive reconstruction fusion module, which maps different sub-band and category features into the same shared embedding space by comparing with category prototypes, reconstructing the feature semantic structure to ensure that the global loss of the fused features converges more easily. Finally, we use a learning mechanism to model the similarity between channel features and use it as the weight of fused sub-band features, thus enhancing the more discriminative features, suppressing the useless features. The experimental accuracy is 79.96% in the BCI competition Ⅳ-Ⅱa dataset. Moreover, the classification effect of sub-band features of different subjects is verified by comparison tests, the category propensity of different sub-band features is verified by confusion matrix tests and the distribution in different classes of each sub-band feature and fused feature are showed by visual analysis, revealing the importance of different sub-band features for the EEG-based MI classification task.

Semantic guided knowledge graph for large-scale zero-shot learning

Zero-shot learning has received growing attention, which aims to improve generalization to unseen concepts. The key challenge in zero-shot tasks is to precisely model the relationship between seen and unseen classes. Most existing zero-shot learning methods capture inter-class relationships via a shared embedding space, leading to inadequate use of relationships and poor performance. Recently, knowledge graph-based methods have emerged as a new trend of zero-shot learning. These methods use a knowledge graph to accurately model the inter-class relationships. However, the currently dominant method for zero-shot learning directly extracts the fixed connection from off-the-shelf WordNet, which will inherit the inherent noise in WordNet. In this paper, we propose a novel method that adopts class-level semantic information as a guidance to construct a new semantic guided knowledge graph (SG-KG), which can correct the errors in the existing knowledge graph and accurately model the inter-class relationships. Specifically, our method includes two main steps: noise suppression and semantic enhancement. Noise suppression is used to eliminate noise edges in the knowledge graph, and semantic enhancement is used to connect two classes with strong relations. To promote high efficient information propagation among classes, we develop a novel multi-granularity fusion network (MGFN) that integrates discriminative information from multiple GCN branches. Extensive experiments on the large-scale ImageNet-21K dataset and AWA2 dataset demonstrate that our method consistently surpasses existing methods and achieves a new state-of-the-art result.

Action-Aware Embedding Enhancement for Image-Text Retrieval

Image-text retrieval plays a central role in bridging vision and language, which aims to reduce the semantic discrepancy between images and texts. Most of existing works rely on refined words and objects representation through the data-oriented method to capture the word-object cooccurrence. Such approaches are prone to ignore the asymmetric action relation between images and texts, that is, the text has explicit action representation (i.e., verb phrase) while the image only contains implicit action information. In this paper, we propose Action-aware Memory-Enhanced embedding (AME) method for image-text retrieval, which aims to emphasize the action information when mapping the images and texts into a shared embedding space. Specifically, we integrate action prediction along with an action-aware memory bank to enrich the image and text features with action-similar text features. The effectiveness of our proposed AME method is verified by comprehensive experimental results on two benchmark datasets.

STonKGs: a sophisticated transformer trained on biomedical text and knowledge graphs.

MotivationThe majority of biomedical knowledge is stored in structured databases or as unstructured text in scientific publications. This vast amount of information has led to numerous machine learning-based biological applications using either text through natural language processing (NLP) or structured data through knowledge graph embedding models. However, representations based on a single modality are inherently limited.ResultsTo generate better representations of biological knowledge, we propose STonKGs, a Sophisticated Transformer trained on biomedical text and Knowledge Graphs (KGs). This multimodal Transformer uses combined input sequences of structured information from KGs and unstructured text data from biomedical literature to learn joint representations in a shared embedding space. First, we pre-trained STonKGs on a knowledge base assembled by the Integrated Network and Dynamical Reasoning Assembler consisting of millions of text-triple pairs extracted from biomedical literature by multiple NLP systems. Then, we benchmarked STonKGs against three baseline models trained on either one of the modalities (i.e. text or KG) across eight different classification tasks, each corresponding to a different biological application. Our results demonstrate that STonKGs outperforms both baselines, especially on the more challenging tasks with respect to the number of classes, improving upon the F1-score of the best baseline by up to 0.084 (i.e. from 0.881 to 0.965). Finally, our pre-trained model as well as the model architecture can be adapted to various other transfer learning applications.Availability and implementationWe make the source code and the Python package of STonKGs available at GitHub (https://github.com/stonkgs/stonkgs) and PyPI (https://pypi.org/project/stonkgs/). The pre-trained STonKGs models and the task-specific classification models are respectively available at https://huggingface.co/stonkgs/stonkgs-150k and https://zenodo.org/communities/stonkgs.Supplementary information Supplementary data are available at Bioinformatics online.

Supervised contrastive learning over prototype-label embeddings for network intrusion detection

Contrastive learning makes it possible to establish similarities between samples by comparing their distances in an intermediate representation space (embedding space) and using loss functions designed to attract/repel similar/dissimilar samples. The distance comparison is based exclusively on the sample features. We propose a novel contrastive learning scheme by including the labels in the same embedding space as the features and performing the distance comparison between features and labels in this shared embedding space. Following this idea, the sample features should be close to its ground-truth (positive) label and away from the other labels (negative labels). This scheme allows to implement a supervised classification based on contrastive learning. Each embedded label will assume the role of a class prototype in embedding space, with sample features that share the label gathering around it. The aim is to separate the label prototypes while minimizing the distance between each prototype and its same-class samples. A novel set of loss functions is proposed with this objective. Loss minimization will drive the allocation of sample features and labels in embedding space. Loss functions and their associated training and prediction architectures are analyzed in detail, along with different strategies for label separation. The proposed scheme drastically reduces the number of pair-wise comparisons, thus improving model performance. In order to further reduce the number of pair-wise comparisons, this initial scheme is extended by replacing the set of negative labels by its best single representative: either the negative label nearest to the sample features or the centroid of the cluster of negative labels. This idea creates a new subset of models which are analyzed in detail.The outputs of the proposed models are the distances (in embedding space) between each sample and the label prototypes. These distances can be used to perform classification (minimum distance label), features dimensionality reduction (using the distances and the embeddings instead of the original features) and data visualization (with 2 or 3D embeddings).Although the proposed models are generic, their application and performance evaluation is done here for network intrusion detection, characterized by noisy and unbalanced labels and a challenging classification of the various types of attacks. Empirical results of the model applied to intrusion detection are presented in detail for two well-known intrusion detection datasets, and a thorough set of classification and clustering performance evaluation metrics are included.

On the Limitations of Visual-Semantic Embedding Networks for Image-to-Text Information Retrieval.

Visual-semantic embedding (VSE) networks create joint image–text representations to map images and texts in a shared embedding space to enable various information retrieval-related tasks, such as image–text retrieval, image captioning, and visual question answering. The most recent state-of-the-art VSE-based networks are: VSE++, SCAN, VSRN, and UNITER. This study evaluates the performance of those VSE networks for the task of image-to-text retrieval and identifies and analyses their strengths and limitations to guide future research on the topic. The experimental results on Flickr30K revealed that the pre-trained network, UNITER, achieved 61.5% on average Recall@5 for the task of retrieving all relevant descriptions. The traditional networks, VSRN, SCAN, and VSE++, achieved 50.3%, 47.1%, and 29.4% on average Recall@5, respectively, for the same task. An additional analysis was performed on image–text pairs from the top 25 worst-performing classes using a subset of the Flickr30K-based dataset to identify the limitations of the performance of the best-performing models, VSRN and UNITER. These limitations are discussed from the perspective of image scenes, image objects, image semantics, and basic functions of neural networks. This paper discusses the strengths and limitations of VSE networks to guide further research into the topic of using VSE networks for cross-modal information retrieval tasks.

Learning product representations for generating reviews for cold products

Existing work in the literature have shown that the number and quality of product ratings and reviews have a direct correlation with the product purchase rates in online e-commerce portals. However, the majority of the products on e-commerce portals do not have any ratings or reviews and are known as cold products (∼90% of products on Amazon are cold). As such, there has been growing interest in generating reviews for cold products by selectively transferring reviews from other similar yet warm products. Our work in this paper focuses on this specific problem and generates reviews for cold products through review selection. Similar to existing work in the literature, our work assumes a relationship between product attribute-values and the reviews that products receive. However, unlike the literature, our method (1) is not restricted to the exact surface form of a product attribute name; and, (2) can distinguish between the same attribute expressed in different forms. We achieve these two important characteristics by proposing methods to learn neural product representations that capture the semantics of product attribute-values as they relate to user reviews. More specifically, our work offers (i) an approach to learn neural representations of product attribute-values within a shared embedding space as product reviews; (ii) a weighted composition strategy to develop product representations from the representation of its attributes; and, (iii) a review selection method that selects relevant reviews for the composed product representation within the neural embedding space. We show through our extensive experiments on five datasets consisting of products from CNET.com and movies from rottentomatoes.com that our method is able to show stronger performance compared to several baselines on ROUGE-2 metrics.

Target-Oriented Deformation of Visual-Semantic Embedding Space

Multimodal embedding is a crucial research topic for cross-modal understanding, data mining, and translation. Many studies have attempted to extract representations from given entities and align them in a shared embedding space. However, because entities in different modalities exhibit different abstraction levels and modality-specific information, it is insufficient to embed related entities close to each other. In this study, we propose the Target-Oriented Deformation Network (TOD-Net), a novel module that continuously deforms the embedding space into a new space under a given condition, thereby adjusting similarities between entities. Unlike methods based on cross-modal attention, TOD-Net is a post-process applied to the embedding space learned by existing embedding systems and improves their performances of retrieval. In particular, when combined with cutting-edge models, TOD-Net gains the state-of-the-art cross-modal retrieval model associated with the MSCOCO dataset. Qualitative analysis reveals that TOD-Net successfully emphasizes entity-specific concepts and retrieves diverse targets via handling higher levels of diversity than existing models.

Differentiable Reasoning on Large Knowledge Bases and Natural Language

Reasoning with knowledge expressed in natural language and Knowledge Bases (KBs) is a major challenge for Artificial Intelligence, with applications in machine reading, dialogue, and question answering. General neural architectures that jointly learn representations and transformations of text are very data-inefficient, and it is hard to analyse their reasoning process. These issues are addressed by end-to-end differentiable reasoning systems such as Neural Theorem Provers (NTPs), although they can only be used with small-scale symbolic KBs. In this paper we first propose Greedy NTPs (GNTPs), an extension to NTPs addressing their complexity and scalability limitations, thus making them applicable to real-world datasets. This result is achieved by dynamically constructing the computation graph of NTPs and including only the most promising proof paths during inference, thus obtaining orders of magnitude more efficient models 1. Then, we propose a novel approach for jointly reasoning over KBs and textual mentions, by embedding logic facts and natural language sentences in a shared embedding space. We show that GNTPs perform on par with NTPs at a fraction of their cost while achieving competitive link prediction results on large datasets, providing explanations for predictions, and inducing interpretable models.

Discriminativeness-Preserved Domain Adaptation for Few-Shot Learning

Existing few-shot learning (FSL) methods make the implicit assumption that the few target class samples are from the same domain as the source class samples. However, this assumption is often invalid in practice - the target classes could come from a different domain. This poses an additional domain adaptation (DA) challenge with few training samples. In this article, the problem of cross-domain few-shot learning (CD-FSL) is approached, which requires solving FSL and DA in a unified framework. To this end, we propose a novel discriminativeness-preserved domain adaptive prototypical network (DPDAPN) model. It is designed to address a specific challenge in CD-FSL: the DA objective means that the source and target data distributions need to be aligned, typically through a shared domain adaptive feature embedding space, but the FSL objective dictates that the target domain per-class distribution must be different from that of any source domain class, meaning aligning the distributions across domains may harm the FSL performance. How to achieve global domain distribution alignment while maintaining source/target per-class discriminativeness thus becomes the key. Our solution is to explicitly enhance the source/target per-class separation before domain adaptive feature embedding learning in DPDAPN to alleviate the negative effect of domain alignment on FSL. Extensive experiments show that our DPDAPN outperforms the state-of-the-art FSL and DA models, as well as their naive combinations.