Variable-length Sequences Research Articles

Comparative analysis and application of soft sensor models in domestic wastewater treatment for advancing sustainability

ABSTRACT This study focuses on the development and evaluation of soft sensor models for predicting NH3-N values in a wastewater treatment process. The study compares the performance of linear regression (LR), neural networks (NN) and random forest regression (RFR) models. The proposed methodology involves optimizing the sequencing batch reactor process using artificial intelligence and an automatic control system. Real-time NH3-N values are obtained by inputting data from electronic conductivity and temperature sensors into the prediction models. Once the predicted NH3-N value falls below the effluent standard, the cycle ends, improving energy efficiency and sustainability by cutting down the agitator and aerator. The research results demonstrate that the RNN-based NH3-N soft sensor built in this study exhibits the best performance, which is promising for wastewater treatment process optimization and evaluation. The results show that sensor model NNR[0.5Y]H exhibits exceptional performance, utilizing recurrent neural network with 5-step input delays. Sensor NNR[0.5Y]H exhibits an R2 of 0.921, an RMSE of 6.110, and an MAE of 4.558. Based on the findings, recurrent neural network (RNN) variants emerge as the most effective modeling technique due to their ability to capture temporal dependencies and handle variable-length sequences. This study provides satisfied performance results for the NNR[0.5Y]H soft sensor model in NH3-N monitoring and process optimization in wastewater treatment, highlighting the effectiveness of recurrent neural networks and their contribution to improving interpretability, accuracy, and adaptability of soft sensor models.

PULMO: Precise utterance-level modeling for speech anti-spoofing

In recent years, most state-of-the-art approaches for spoofed speech detection have been based on convolutional neural networks (CNNs). Most neural networks, including CNNs, are trained in minibatch units, where all input data in each minibatch must have the same shape. Therefore, for minibatch training, each utterance is first either padded or truncated because utterances are variable-length sequences and thus cannot be directly fed into networks in minibatch units. However, modeling either a padded or truncated utterance, rather than the original one, makes it unfeasible to capture the entire context as is: padding could propagate even unwanted information, like artifacts, in the original utterance, and truncation inevitably loses some information. With these information distortions, model could get stuck in a suboptimal solution. To fill this gap, we proposeÚ a method for precise utterance-level modeling that enables minibatch-wise utterance-level modeling of variable-length utterances while minimizing the information distortions. The proposed method comprises sequence segmentation followed by segment aggregation. Sequence segmentation feeds variable-length utterances in the minibatch unit by decomposing each of them into fixed-length segments, which enables parallel processing of variable-length utterances without the uncertainty in input length. Segment aggregation plays a role in aggregating the segment embeddings by utterance to encode the entire information of each utterance. The experimental results of the evaluation trials of ASVspoof 2019 and 2021 indicate that the proposed method shows up to 84.9 % and 97.6 % relative equal error rate reductions on logical and physical access scenarios, respectively. Furthermore, the proposed method reduced the FLOPs for an epoch by 6 %.

Design of an efficient Transformer-XL model for enhanced pseudo code to Python code conversion

The landscape of programming has long been challenged by the task of transforming pseudo code into executable Python code, a process traditionally marred by its labor-intensive nature and the necessity for a deep understanding of both logical frameworks and programming languages. Existing methodologies often grapple with limitations in handling variable-length sequences and maintaining context over extended textual data. Addressing these challenges, this study introduces an innovative approach utilizing the Transformer-XL model, a significant advancement in the domain of deep learning. The Transformer-XL architecture, an evolution of the standard Transformer, adeptly processes variable-length sequences and captures extensive contextual dependencies, thereby surpassing its predecessors in handling natural language processing (NLP) and code synthesis tasks. The proposed model employs a comprehensive process involving data preprocessing, model input encoding, a self-attention mechanism, contextual encoding, language modeling, and a meticulous decoding process, followed by post-processing. The implications of this work are far-reaching, offering a substantial leap in the automation of code conversion. As the field of NLP and deep learning continues to evolve, the Transformer-XL based model is poised to become an indispensable tool in the realm of programming, setting a new benchmark for automated code synthesis.

Enhancing handwritten text recognition accuracy with gated mechanisms

Handwritten Text Recognition (HTR) is a challenging task due to the complex structures and variations present in handwritten text. In recent years, the application of gated mechanisms, such as Long Short-Term Memory (LSTM) networks, has brought significant advancements to HTR systems. This paper presents an overview of HTR using a gated mechanism and highlights its novelty and advantages. The gated mechanism enables the model to capture long-term dependencies, retain relevant context, handle variable length sequences, mitigate error propagation, and adapt to contextual variations. The pipeline involves preprocessing the handwritten text images, extracting features, modeling the sequential dependencies using the gated mechanism, and decoding the output into readable text. The training process utilizes annotated datasets and optimization techniques to minimize transcription discrepancies. HTR using a gated mechanism has found applications in digitizing historical documents, automatic form processing, and real-time transcription. The results show improved accuracy and robustness compared to traditional HTR approaches. The advancements in HTR using a gated mechanism open up new possibilities for effectively recognizing and transcribing handwritten text in various domains. This research does a better job than the most recent iteration of the HTR system when compared to five different handwritten datasets (Washington, Saint Gall, RIMES, Bentham and IAM). Smartphones and robots are examples of low-cost computing devices that can benefit from this research.

LipNet: End-to-End Lipreading

Lipreading is the task of decoding text from the movement of a speaker’s mouth. This research presents the development of an advanced end-to-end lipreading system. Leveraging deep learning architectures and multimodal fusion techniques, the proposed system interprets spoken language solely from visual cues, such as lip movements. Through meticulous data collection, annotation, preprocessing, model development, and evaluation, diverse datasets encompassing various speakers, accents, languages, and environmental conditions are curated to ensure robustness and generalization. Conventional methods divided the task into two phases: prediction and designing or learning visual characteristics. Most deep lipreading methods are trainable from end to end. In the past, lipreading has been tackled using tedious and sometimes unsatisfactory techniques that break down speech into smaller units like phonemes or visemes. But these methods often fail when faced with real-world problems, such contextual factors, accents, and differences in speech patterns. Nevertheless, current research on end-to-end trained models only carries out word classification; sentence-level sequence prediction is not included. LipNet is an end-to-end trained model that uses spatiotemporal convolutions, a recurrent network, and the connectionist temporal classification loss to translate a variable-length sequence of video frames to text. LipNet breaks from this traditional paradigm by using an all-encompassing, end-to-end approach supported by deep learning algorithms, Convolutional neural networks (CNNs) and recurrent neural networks (RNNs), which are skilled at processing sequential data and extracting high-level representations, are fundamental to LipNet's architecture.LipNet achieves 95.2% accuracy in sentence-level on the GRID corpus, overlapped speaker split task, outperforming experienced human lipreaders and the previous 86.4% word-level state-of-the-art accuracy.The results underscore the transformative potential of the lipreading system in real-world applications, particularly in domains such as assistive technology and human-computer interaction, where it can significantly improve communication accessibility and inclusivity for individuals with hearing impairments.

Deephullnet: a deep learning approach for solving the convex hull and concave hull problems with transformer

ABSTRACT Convex and concave hulls originating from computational geometry are widely applied in practice. For instance, to determine the boundaries of a geographical area within a group of cities, convex hulls can represent the approximate boundaries of the areas. Concave hulls can accurately describe the shape of the area. The traditional methods for solving two-dimensional convex hull problems include the Jarvis March algorithm and the Graham scanning algorithm. The K-nearest neighbours and alpha-shape methods are designed for solving the concave hull problem. Other than traditional algorithms, we consider using deep neural networks to handle the convex and concave hull problems. General neural networks cannot deal with the problem whose output is a variable-length sequence. Our study provides a machine-learning approach for solving the convex hull and concave hull problems. We combine the Pointer network (Ptr-Net) with the Transformer model and propose the DeepHullNet. The trained DeepHullNet can provide a feasible solution in the majority of cases. The experimental results show that DeepHullNet outperforms the original Ptr-Net regarding accuracy. Compared with traditional methods, DeepHullNet can generate a good solution quickly, and the running time is shortened by nearly 100 times based on GPU.

Efficient Production of Enterovirus 71 (EV71) Virus-like Particles by Controlling Promoter Strength in Insect Cells.

This study was conducted to efficiently produce virus-like particles (VLPs) of enterovirus 71 (EV71), a causative virus of hand, foot, and mouth disease (HFMD). The expression level of the P1 precursor, a structural protein of EV71, was modified to increase VLP production, and the optimal expression level and duration of the 3CD protein for P1 cleavage were determined. The expression level and duration of 3CD were controlled by the p10 promoter, which was weakened by repeated burst sequence (BS) applications, as well as the OpIE2 promoter, which was weakened by the insertion of random untranslated region sequences of various lengths. The cleavage and production efficiency of the P1 precursor were compared based on the expression time and level of 3CD, revealing that the p10-BS5 promoter with four repeated BSs was the most effective. When P1 and 3CD were expressed using the hyperexpression vector and the p10-BS5 promoter, high levels of structural protein production and normal HFMD-VLP formation were observed, respectively. This study suggests that the production efficiency of HFMD-VLPs can be significantly enhanced by increasing the expression of the P1 precursor and controlling the amount and duration of 3CD expression.

ARGNet: using deep neural networks for robust identification and classification of antibiotic resistance genes from sequences

BackgroundEmergence of antibiotic resistance in bacteria is an important threat to global health. Antibiotic resistance genes (ARGs) are some of the key components to define bacterial resistance and their spread in different environments. Identification of ARGs, particularly from high-throughput sequencing data of the specimens, is the state-of-the-art method for comprehensively monitoring their spread and evolution. Current computational methods to identify ARGs mainly rely on alignment-based sequence similarities with known ARGs. Such approaches are limited by choice of reference databases and may potentially miss novel ARGs. The similarity thresholds are usually simple and could not accommodate variations across different gene families and regions. It is also difficult to scale up when sequence data are increasing.ResultsIn this study, we developed ARGNet, a deep neural network that incorporates an unsupervised learning autoencoder model to identify ARGs and a multiclass classification convolutional neural network to classify ARGs that do not depend on sequence alignment. This approach enables a more efficient discovery of both known and novel ARGs. ARGNet accepts both amino acid and nucleotide sequences of variable lengths, from partial (30–50 aa; 100–150 nt) sequences to full-length protein or genes, allowing its application in both target sequencing and metagenomic sequencing. Our performance evaluation showed that ARGNet outperformed other deep learning models including DeepARG and HMD-ARG in most of the application scenarios especially quasi-negative test and the analysis of prediction consistency with phylogenetic tree. ARGNet has a reduced inference runtime by up to 57% relative to DeepARG.ConclusionsARGNet is flexible, efficient, and accurate at predicting a broad range of ARGs from the sequencing data. ARGNet is freely available at https://github.com/id-bioinfo/ARGNet, with an online service provided at https://ARGNet.hku.hk.7nR91jX7ySRMRPZrE5tsTyVideo

The onset of instability and bifurcations in the transitional wake of two tandem square cylinders

Unsteady three-dimensional simulations are performed to elucidate the hidden transitional flow dynamics and Hopf bifurcations along the topological corelines of the created von Kármán streets behind a pair of square cylinders positioned in tandem. Our simulations provide significant new insight into the three-dimensional wake evolution and governing flow physics. We explain how pressure, velocity, and vorticity fluctuation along the Kármán vortex corelines in the increasingly unstable wake amplify, facilitating the growth of various modal instability patterns. The existing knowledge of wake transition through the intertwining of modes A, B, and C instabilities and associated linear stability analysis helped to gain some insight into the overall wake feature. The current study explains how exactly the transitional disturbances physically spread behind a pair of inline and tandem cylinders through the self-excited spanwise-periodic oscillation of the wake and created a sequence of variable length scaled Hopf bifurcations and their swapping for varied gaps between the cylinders and Reynolds number. The growth of a slow mode of the spectral frequency at the bifurcation points seemed crucial in initiating the near-transitional flow irregularity.

A Two-Layer Self-Organizing Map with Vector Symbolic Architecture for Spatiotemporal Sequence Learning and Prediction.

We propose a new nature- and neuro-science-inspired algorithm for spatiotemporal learning and prediction based on sequential recall and vector symbolic architecture. A key novelty is the learning of spatial and temporal patterns as decoupled concepts where the temporal pattern sequences are constructed using the learned spatial patterns as an alphabet of elements. The decoupling, motivated by cognitive neuroscience research, provides the flexibility for fast and adaptive learning with dynamic changes to data and concept drift and as such is better suited for real-time learning and prediction. The algorithm further addresses several key computational requirements for predicting the next occurrences based on real-life spatiotemporal data, which have been found to be challenging with current state-of-the-art algorithms. Firstly, spatial and temporal patterns are detected using unsupervised learning from unlabeled data streams in changing environments; secondly, vector symbolic architecture (VSA) is used to manage variable-length sequences; and thirdly, hyper dimensional (HD) computing-based associative memory is used to facilitate the continuous prediction of the next occurrences in sequential patterns. The algorithm has been empirically evaluated using two benchmark and three time-series datasets to demonstrate its advantages compared to the state-of-the-art in spatiotemporal unsupervised sequence learning where the proposed ST-SOM algorithm is able to achieve 45% error reduction compared to HTM algorithm.

EndoGenius: Optimized Neuropeptide Identification from Mass Spectrometry Datasets.

Neuropeptides represent a unique class of signaling molecules that have garnered much attention but require special consideration when identifications are gleaned from mass spectra. With highly variable sequence lengths, neuropeptides must be analyzed in their endogenous state. Further, neuropeptides share great homology within families, differing by as little as a single amino acid residue, complicating even routine analyses and necessitating optimized computational strategies for confident and accurate identifications. We present EndoGenius, a database searching strategy designed specifically for elucidating neuropeptide identifications from mass spectra by leveraging optimized peptide-spectrum matching approaches, an expansive motif database, and a novel scoring algorithm to achieve broader representation of the neuropeptidome and minimize reidentification. This work describes an algorithm capable of reporting more neuropeptide identifications at 1% false-discovery rate than alternative software in five Callinectes sapidus neuronal tissue types.

Genome-Wide Studies of FH Family Members in Soybean (Glycine max) and Their Responses under Abiotic Stresses.

Formins or formin homology 2 (FH2) proteins, evolutionarily conserved multi-domain proteins in eukaryotes, serve as pivotal actin organizers, orchestrating the structure and dynamics of the actin cytoskeleton. However, a comprehensive investigation into the formin family and their plausible involvement in abiotic stress remains undocumented in soybean (Glycine max). In the current study, 34 soybean FH (GmFH)family members were discerned, their genomic distribution spanning the twenty chromosomes in a non-uniform pattern. Evolutionary analysis of the FH gene family across plant species delineated five discernible groups (Group I to V) and displayed a closer evolutionary relationship within Glycine soja, Glycine max, and Arabidopsis thaliana. Analysis of the gene structure of GmFH unveiled variable sequence lengths and substantial diversity in conserved motifs. Structural prediction in the promoter regions of GmFH gene suggested a large set of cis-acting elements associated with hormone signaling, plant growth and development, and stress responses. The investigation of the syntenic relationship revealed a greater convergence of GmFH genes with dicots, indicating a close evolutionary affinity. Transcriptome data unveiled distinctive expression patterns of several GmFH genes across diverse plant tissues and developmental stages, underscoring a spatiotemporal regulatory framework governing the transcriptional dynamics of GmFH gene. Gene expression and qRT-PCR analysis identified many GmFH genes with a dynamic pattern in response to abiotic stresses, revealing their potential roles in regulating plant stress adaptation. Additionally, protein interaction analysis highlighted an intricate web of interactions among diverse GmFH proteins. These findings collectively underscore a novel biological function of GmFH proteins in facilitating stress adaptation in soybeans.

Characterization of Fowlpox Virus.

The complex cytoplasmic DNA virus known as the fowlpox virus (FWPV) is a member of the avipoxvirus genus, Subfamily Chordopoxvirinae, and Family Poxviridae. The large genome size of FWPV makes it a potential vector for the creation of vaccines against a range of serious veterinary and human ailments. It also allows for multiple gene insertion and the generation of abortive infection in mammalian cells. The virus, which causes fowlpox in chickens and turkeys, is mainly transmitted to poultry through aerosols or biting insects. Fowlpox is a highly contagious disease that affects both domestic and wild birds, causing cutaneous and/or diphtheritic illnesses. To control the illness, strict hygiene practices and immunization with FWPV attenuated strains or antigenically similar pigeon pox virus vaccines are employed. Recent years have seen an increase in fowlpox outbreaks in chicken flocks, primarily due to the introduction of novel forms of FWPV. It is believed that the pathogenic characteristics of these strains are enhanced by the integration of reticuloendotheliosis virus sequences of variable lengths into the FWPV genome. The standard laboratory diagnosis of FPV involves histopathological analysis, electron microscopy, virus isolation on chorioallantoic membrane (CAM) of embryonated chicken eggs or cell cultures, and serologic techniques. For quick and consistent diagnosis, polymerase chain reaction (PCR) has proven to be the most sensitive method. PCR is used in concert with restriction endonuclease enzyme analysis (REA) to identify, differentiate, and characterize the molecular makeup of isolates of the fowlpox virus. Sequencing of the amplified fragments is then done.

A hybrid Bi-LSTM and RBM approach for advanced underwater object detection.

This research addresses the imperative need for efficient underwater exploration in the domain of deep-sea resource development, highlighting the importance of autonomous operations to mitigate the challenges posed by high-stress underwater environments. The proposed approach introduces a hybrid model for Underwater Object Detection (UOD), combining Bi-directional Long Short-Term Memory (Bi-LSTM) with a Restricted Boltzmann Machine (RBM). Bi-LSTM excels at capturing long-term dependencies and processing sequences bidirectionally to enhance comprehension of both past and future contexts. The model benefits from effective feature learning, aided by RBMs that enable the extraction of hierarchical and abstract representations. Additionally, this architecture handles variable-length sequences, mitigates the vanishing gradient problem, and achieves enhanced significance by capturing complex patterns in the data. Comprehensive evaluations on brackish, and URPC 2020 datasets demonstrate superior performance, with the BiLSTM-RBM model showcasing notable accuracies, such as big fish 98.5 for the big fish object in the brackish dataset and 98 for the star fish object in the URPC dataset. Overall, these findings underscore the BiLSTM-RBM model's suitability for UOD, positioning it as a robust solution for effective underwater object detection in challenging underwater environments.

HostNet: improved sequence representation in deep neural networks for virus-host prediction

BackgroundThe escalation of viruses over the past decade has highlighted the need to determine their respective hosts, particularly for emerging ones that pose a potential menace to the welfare of both human and animal life. Yet, the traditional means of ascertaining the host range of viruses, which involves field surveillance and laboratory experiments, is a laborious and demanding undertaking. A computational tool with the capability to reliably predict host ranges for novel viruses can provide timely responses in the prevention and control of emerging infectious diseases. The intricate nature of viral-host prediction involves issues such as data imbalance and deficiency. Therefore, developing highly accurate computational tools capable of predicting virus-host associations is a challenging and pressing demand.ResultsTo overcome the challenges of virus-host prediction, we present HostNet, a deep learning framework that utilizes a Transformer-CNN-BiGRU architecture and two enhanced sequence representation modules. The first module, k-mer to vector, pre-trains a background vector representation of k-mers from a broad range of virus sequences to address the issue of data deficiency. The second module, an adaptive sliding window, truncates virus sequences of various lengths to create a uniform number of informative and distinct samples for each sequence to address the issue of data imbalance. We assess HostNet's performance on a benchmark dataset of “Rabies lyssavirus” and an in-house dataset of “Flavivirus”. Our results show that HostNet surpasses the state-of-the-art deep learning-based method in host-prediction accuracies and F1 score. The enhanced sequence representation modules, significantly improve HostNet's training generalization, performance in challenging classes, and stability.ConclusionHostNet is a promising framework for predicting virus hosts from genomic sequences, addressing challenges posed by sparse and varying-length virus sequence data. Our results demonstrate its potential as a valuable tool for virus-host prediction in various biological contexts. Virus-host prediction based on genomic sequences using deep neural networks is a promising approach to identifying their potential hosts accurately and efficiently, with significant impacts on public health, disease prevention, and vaccine development.

Temporal convolutional approach with residual multi-head attention mechanism for remaining useful life of manufacturing tools

Predicting the remaining useful life of manufacturing tools based on condition maintenance is a critical task in prognostics and health management systems. Existing methods have limitations in modeling variable-length sequence data and extracting sufficient local and global dependencies. To address these issues, a temporal convolutional approach with residual multi-head attention mechanism is proposed. The temporal convolution efficiently models variable-length sequence data and extracts long-term dependency features in the time dimension. The residual multi-head attention module captures local and global dependency features in the sequence features and extracts the tool key degradation features, which further improves the performance of the model. To validate the predictive performance of the temporal convolutional approach with residual multi-head attention mechanism, two manufacturing tool datasets are used in the remaining useful life experiments. The results demonstrate the effectiveness and superiority of the proposed method compared to other advanced methods.

ProtWave-VAE: Integrating Autoregressive Sampling with Latent-Based Inference for Data-Driven Protein Design.

Deep generative models (DGMs) have shown great success in the understanding and data-driven design of proteins. Variational autoencoders (VAEs) are a popular DGM approach that can learn the correlated patterns of amino acid mutations within a multiple sequence alignment (MSA) of protein sequences and distill this information into a low-dimensional latent space to expose phylogenetic and functional relationships and guide generative protein design. Autoregressive (AR) models are another popular DGM approach that typically lacks a low-dimensional latent embedding but does not require training sequences to be aligned into an MSA and enable the design of variable length proteins. In this work, we propose ProtWave-VAE as a novel and lightweight DGM, employing an information maximizing VAE with a dilated convolution encoder and an autoregressive WaveNet decoder. This architecture blends the strengths of the VAE and AR paradigms in enabling training over unaligned sequence data and the conditional generative design of variable length sequences from an interpretable, low-dimensional learned latent space. We evaluated the model's ability to infer patterns and design rules within alignment-free homologous protein family sequences and to design novel synthetic proteins in four diverse protein families. We show that our model can infer meaningful functional and phylogenetic embeddings within latent spaces and make highly accurate predictions within semisupervised downstream fitness prediction tasks. In an application to the C-terminal SH3 domain in the Sho1 transmembrane osmosensing receptor in baker's yeast, we subject ProtWave-VAE-designed sequences to experimental gene synthesis and select-seq assays for the osmosensing function to show that the model enables synthetic protein design, conditional C-terminus diversification, and engineering of the osmosensing function into SH3 paralogues.

Generation of random sequences using DNA cryptography for OTP encryption

DNA cryptography is considered as a promising new field for improving cybersecurity and privacy using parallelism and storage capacity of DNA molecules. In this work, novel technique is proposed for generating random and unpredictable keys for symmetric One Time Pad cryptosystems. To achieve this, publicly available DNA sequences stored in genetic databases are used as a source of inherently random, diverse, and variable-length sequences. The proposed approach is based on two steps. The first step uses a DNA self-assembly structure, mathematical operations and DNA techniques. In the second step, an entropy source and chaotic function are incorporated as a random number generator and pseudorandom number generator, respectively, to further enhance the randomness and unpredictability of the generated OTP keys. Additionally, a secure key exchange approach is also proposed including minimization and optimization in both secure and public channels used for key exchange.

Protein-DNA Binding Residues Prediction Using a Deep Learning Model With Hierarchical Feature Extraction.

Biologically important effects occur when proteins bind to other substances, of which binding to DNA is a crucial one. Therefore, accurate identification of protein-DNA binding residues is important for further understanding of the protein-DNA interaction mechanism. Although wet-lab methods can accurately obtain the location of bound residues, it requires significant human, financial and time costs. There is thus an urgent need to develop efficient computational-based methods. Most current state-of-the-art methods are two-step approaches: the first step uses a sliding window technique to extract residue features; the second step uses each residue as an input to the model for prediction. This has a negative impact on the efficiency of prediction and ease of use. In this study, we propose a sequence-to-sequence (seq2seq) model that can input the entire protein sequence of variable length and use two modules, Transformer Encoder Block and Feature Extracting Block, for hierarchical feature extraction, where Transformer Encoder Block is used to extract global features, and then Feature Extracting Block is used to extract local features to further improve the recognition capability of the model. The comparison results on two benchmark datasets, namely PDNA-543 and PDNA-41, prove the effectiveness of our method in identifying protein-DNA binding residues.

Variable-length sequence model for attribute detection in the image

Holistic scene understanding is a challenging problem in computer vision. Most recent researches in this field were focusing on the object detection, the semantic segmentation and the relationship detection tasks. The attribute can provide meaningful information for the object instance, thus the object instance can be expressed more detail in the scene understanding. However, most researches in this field have been limited to several special conditions. Such as, several researches were just focusing on the attribute of special object class, because their solutions were aimed at a limited-scenarios, their methods are hardly to generalize in other scenarios. We also find that most of the research for multi-attribute detection task were only regarding each attribute as binary class and simply use the multi-binary-classifier method for the attribute detection. But these strategies above not consider the relation between each pair of the attributes, they will fall into trouble in the “imperfect” attribute dataset (which is labeled with the missing and incomplete annotations), and they will have low performance in the long-tail attribute class (which has lower rank of annotation and more missing labels). In this paper, we focus on the multi-attribute detection for a variant of object classes and take the relation between attributes into consideration. We propose a GRU-based model to detect a variable-length attribute sequence with a customized loss compute method to solve the “imperfect” attribute dataset problem. Furthermore, we perform ablative studies to prove the effectiveness of each part of our method. Finally, we compare our model with several existed multi-attribute detection methods on VG (Visual Genome) and CUB200 bird datasets to prove the superior performance of the proposed model.