Hierarchical Semantic Compression for Consistent Image Semantic Restoration.

In today’s big data era, a key requirement is to implement intelligent semantic analysis (such as image recognition) on data gathered from an extensive array of smart devices in Artificial Intelligence IoT (AIoT) scenarios, all of which is processed at central cloud service providers. Recent advancements in deep-learning-based image compression have fostered semantic compression between machines. However, the deployment of an overparameterized encoder on Internet of Things (IoT) devices remains a challenge due to their restricted computing and storage capabilities. To tackle this issue, we propose a novel approach named compressed sensing (CS)-based asymmetric semantic image compression (CS-ASIC), explicitly designed for resource-constrained AIoT systems. This asymmetric semantic compression scheme intends to surpass the limitations of IoT devices, thereby facilitating efficient semantic compression for machine vision tasks. CS-ASIC notably includes a lightweight front encoder founded on deep image CS techniques, which utilizes rich image priors to learn measurement matrices for sampling. In tandem, a deep iterative decoder is designed cooperatively with the linear encoder offloaded at the server to enhance image reconstruction and semantic analysis across various semantic analysis tasks. Furthermore, we introduce a groundbreaking lossy CS semantic rate-distortion theoretical framework that justifies a compromise in rate for extended semantic distortion. Extensive experimental results underscore the superiority of the proposed CS-ASIC concerning the signal-semantic rate-distortion tradeoff, and its lower encoding complexity over existing codecs in an AIoT simulation environment.

Despite the recent impressive breakthroughs in text-to-image generation, generative models have difficulty in capturing the data distribution of underrepresented attribute compositions while over-memorizing overrepresented attribute compositions, which raises public concerns about their robustness and fairness. To tackle this challenge, we propose ACTIG, an attribute-centric compositional text-to-image generation framework. We present an attribute-centric feature augmentation and a novel image-free training scheme, which greatly improves model’s ability to generate images with underrepresented attributes. We further propose an attribute-centric contrastive loss to avoid overfitting to overrepresented attribute compositions. We validate our framework on the CelebA-HQ and CUB datasets. Extensive experiments show that the compositional generalization of ACTIG is outstanding, and our framework outperforms previous works in terms of image quality and text-image consistency. The source code and trained models are publicly available at https://github.com/yrcong/ACTIG.

In this paper, we address the task of semantic-guided scene generation. One open challenge widely observed in global image-level generation methods is the difficulty of generating small objects and detailed local texture. To tackle this issue, in this work we consider learning the scene generation in a local context, and correspondingly design a local class-specific generative network with semantic maps as a guidance, which separately constructs and learns sub-generators concentrating on the generation of different classes, and is able to provide more scene details. To learn more discriminative class-specific feature representations for the local generation, a novel classification module is also proposed. To combine the advantage of both global image-level and the local class-specific generation, a joint generation network is designed with an attention fusion module and a dual-discriminator structure embedded. Extensive experiments on two scene image generation tasks show superior generation performance of the proposed model. State-of-the-art results are established by large margins on both tasks and on challenging public benchmarks. The source code and trained models are available at https://github.com/Ha0Tang/LGGAN.

Statistical language modeling techniques have successfully been applied to large source code corpora, yielding a variety of new software development tools, such as tools for code suggestion, improving readability, and API migration. A major issue with these techniques is that code introduces new vocabulary at a far higher rate than natural language, as new identifier names proliferate. Both large vocabularies and out-of-vocabulary issues severely affect Neural Language Models (NLMs) of source code, degrading their performance and rendering them unable to scale. In this paper, we address this issue by: 1) studying how various modelling choices impact the resulting vocabulary on a large-scale corpus of 13,362 projects; 2) presenting an open vocabulary source code NLM that can scale to such a corpus, 100 times larger than in previous work; and 3) showing that such models outperform the state of the art on three distinct code corpora (Java, C, Python). To our knowledge, these are the largest NLMs for code that have been reported. All datasets, code, and trained models used in this work are publicly available.

It is hard to generate an image at target view well for previous cross-view image translation methods that directly adopt a simple encoder-decoder or U-Net structure, especially for drastically different views and severe deformation cases. To ease this problem, we propose a novel two-stage framework with a new Cascaded Cross MLP-Mixer (CrossMLP) sub-network in the first stage and one refined pixel-level loss in the second stage. In the first stage, the CrossMLP sub-network learns the latent transformation cues between image code and semantic map code via our novel CrossMLP blocks. Then the coarse results are generated progressively under the guidance of those cues. Moreover, in the second stage, we design a refined pixel-level loss that eases the noisy semantic label problem with more reasonable regularization in a more compact fashion for better optimization. Extensive experimental results on Dayton~\cite{vo2016localizing} and CVUSA~\cite{workman2015wide} datasets show that our method can generate significantly better results than state-of-the-art methods. The source code and trained models are available at https://github.com/Amazingren/CrossMLP.

Human leukocyte antigen (HLA) complex molecules play an essential role in immune interactions by presenting peptides on the cell surface to T cells. With significant deep learning progress, a series of neural network-based models have been proposed and demonstrated with their excellent performances for peptide-HLA class I binding prediction. However, there is still a lack of effective binding prediction models for HLA class II protein binding with peptides due to its inherent challenges. We present a novel sequence-based pan-specific neural network structure, DeepSeaPanII, for peptide-HLA class II binding prediction in this work. Our model is an end-to-end neural network model without the need for pre-or post-processing on input samples compared with existing pan-specific models. Besides state-of-the-art performance in binding affinity prediction, DeepSeqPanII can also extract biological insight on the binding mechanism over the peptide by its attention mechanism-based binding core prediction capability. The leave-one-allele-out cross-validation and benchmark evaluation results show that our proposed network model achieved state-of-the-art performance in HLA-II peptide binding. The source code and trained models are freely available at https://github.com/pcpLiu/DeepSeqPanII.

Siamese trackers demonstrated high performance in object tracking due to their balance between accuracy and speed. Unlike classification-based CNNs, deep similarity networks are specifically designed to address the image similarity problem and thus are inherently more appropriate for the tracking task. However, Siamese trackers mainly use the last convolutional layers for similarity analysis and target search, which restricts their performance. In this paper, we argue that using a single convolutional layer as feature representation is not an optimal choice in a deep similarity framework. We present a Multiple Features-Siamese Tracker (MFST), a novel tracking algorithm exploiting several hierarchical feature maps for robust tracking. Since convolutional layers provide several abstraction levels in characterizing an object, fusing hierarchical features allows to obtain a richer and more efficient representation of the target. Moreover, we handle the target appearance variations by calibrating the deep features extracted from two different CNN models. Based on this advanced feature representation, our method achieves high tracking accuracy, while outperforming the standard siamese tracker on object tracking benchmarks.The source code and trained models are available at https://github.com/zhenxili96/MFST.

In wheat breeding, spike number is a key indicator for evaluating wheat yield, and the timely and accurate acquisition of wheat spike number is of great practical significance for yield prediction. In actual production; the method of using an artificial field survey to count wheat spikes is time-consuming and labor-intensive. Therefore, this paper proposes a method based on YOLOv5s with an improved attention mechanism, which can accurately detect the number of small-scale wheat spikes and better solve the problems of occlusion and cross-overlapping of the wheat spikes. This method introduces an efficient channel attention module (ECA) in the C3 module of the backbone structure of the YOLOv5s network model; at the same time, the global attention mechanism module (GAM) is inserted between the neck structure and the head structure; the attention mechanism can be more Effectively extract feature information and suppress useless information. The result shows that the accuracy of the improved YOLOv5s model reached 71.61% in the task of wheat spike number, which was 4.95% higher than that of the standard YOLOv5s model and had higher counting accuracy. The improved YOLOv5s and YOLOv5m have similar parameters, while RMSE and MEA are reduced by 7.62 and 6.47, respectively, and the performance is better than YOLOv5l. Therefore, the improved YOLOv5s method improves its applicability in complex field environments and provides a technical reference for the automatic identification of wheat spike numbers and yield estimation. Labeled images, source code, and trained models are available at: https://github.com/228384274/improved-yolov5.

Style may refer to different concepts (e.g. painting style, hairstyle, texture, color, filter, etc.) depending on how the feature space is formed. In this work, we propose a novel idea of interpreting the lighting in the single- and multi-illuminant scenes as the concept of style. To verify this idea, we introduce an enhanced auto white-balance (AWB) method that models the lighting in single- and mixed-illuminant scenes as the style factor. Our AWB method does not require any illumination estimation step, yet contains a network learning to generate the weighting maps of the images with different WB settings. Proposed network utilizes the style information, extracted from the scene by a multi-head style extraction module. AWB correction is completed after blending these weighting maps and the scene. Experiments on single- and mixed-illuminant datasets demonstrate that our proposed method achieves promising correction results when compared to the recent works. This shows that the lighting in the scenes with multiple illuminations can be modeled by the concept of style. Source code and trained models are available on https://github.com/birdortyedi/lighting-as-style-awb-correction.

Facial analysis is an active research area in computer vision, with many practical applications. Most of the existing studies focus on addressing one specific task and maximizing its performance. For a complete facial analysis system, one needs to solve these tasks efficiently to ensure a smooth experience. In this work, we present a system-level design of a real-time facial analysis system. With a collection of deep neural networks for object detection, classification, and regression, the system recognizes age, gender, facial expression, and facial similarity for each person that appears in the camera view. We investigate the parallelization and interplay of individual tasks. Results on common off-the-shelf architecture show that the system’s accuracy is comparable to the state-of-the-art methods, and the recognition speed satisfies real-time requirements. Moreover, we propose a multitask network for jointly predicting the first three attributes, i.e., age, gender, and facial expression. Source code and trained models are available at https://github.com/mahehu/TUT-live-age-estimator.

SoybeanTracer: An In-Field scene property-based framework for high-throughput soybean canopy coverage extraction and evaluation

Most modern satellites can provide two types of images: 1) panchromatic (PAN) image and 2) multispectral (MS) image. The former has high spatial resolution and low spectral resolution, while the latter has high spectral resolution and low spatial resolution. To obtain images with both high spectral and spatial resolution, pansharpening has emerged to fuse the spatial information of the PAN image and the spectral information of the MS image. However, most pansharpening methods fail to preserve spatial and spectral information simultaneously. In this article, we propose a framelet-based convolutional neural network (CNN) for pansharpening which makes it possible to pursue both high spectral and high spatial resolution. Our network consists of three subnetworks: 1) feature embedding net; 2) feature fusion net; and 3) framelet prediction net. Different from conventional CNN methods directly inferring high-resolution MS images, our approach learns to predict their framelet coefficients from available PAN and MS images. The introduction of multilevel feature aggregation and hybrid residual connection makes full use of spatial information of PAN image and spectral information of MS image. Quantitative and qualitative experiments at reduced- and full-resolution demonstrate that the proposed method achieves more appealing results than other state-of-the-art pansharpening methods. The source code and trained models are available at https://github.com/TingMAC/FrMLNet.

Convolution-based methods are increasingly being used in medical image segmentation tasks and have shown good performance, but there are always problems in segmenting edge parts. These methods all have the following challenges: 1) Previous methods do not highlight the relationship between foreground and background in segmented regions, which is helpful for complex segmentation edges, 2) inductive bias of the convolutional layer leads to the fact that the extracted information is mainly the main part of the segmented area, and cannot effectively perceive complex edge changes and the aggregation of small and many segmented areas,3) different regions around the segmentation edge have different reference values for segmentation, and the ordering of these values is more important when the segmentation task is more complex. To address these challenges, we propose the CM-MLP framework on Multi-scale Feature Interaction (MFI) block and Axial Context Relation Encoder (ACRE) block for accurate segmentation of the edge of medical image. In the MFI block, we propose the Cascade Multi-scale MLP (Cascade MLP) to process all local information from the deeper layers of the network simultaneously, using Squeeze and Excitation in Space(SES) to process and redistribute the weights of all windows in Cascade MLP and utilize a cascade multi-scale mechanism to fuse discrete local information gradually. Then, multiple ACRE blocks cooperate with the deep supervision mechanism to gradually explore the boundary relationship between the foreground and the background, and gradually fine-tune the edges of the medical image. The segmentation accuracy (Dice) of our proposed CM-MLP framework reaches 96.98%, 96.67%, and 83.83% on three benchmark datasets: CVC-ClinicDB dataset, sub-Kvasir dataset, and our in-house dataset, respectively, which significantly outperform the state-of-the-art method. The source code and trained models will be available at https://github.com/ProgrammerHyy/CM-MLP.

We present a proof of concept for mining JWST imaging data for anomalous galaxy populations using a conditional Generative Adversarial Network (cGAN). We train our model to predict long wavelength NIRcam fluxes (LW: F277W, F356W, F444W between 2.4 and 5.0 μm) from short wavelength fluxes (SW: F115W, F150W, F200W between 0.6 and 2.3 μm) in ∼2000 galaxies. We test the cGAN on a population of 37 Extremely Red Objects (EROs) discovered by the CEERS JWST Team. Despite their red long wavelength colors, the EROs have blue short wavelength colors (F150W − F200W ∼ 0 mag) indicative of bimodal SEDs. Surprisingly, given their unusual SEDs, we find that the cGAN accurately predicts the LW NIRcam fluxes of the EROs. However, it fails to predict LW fluxes for other rare astronomical objects, such as a merger between two galaxies, suggesting that the cGAN can be used to detect some anomalies. (The source code and trained model (Pearce-Casey ) are available at https://github.com/RubyPC/Anomaly_Detection_with_cGANs.)

It is often valuable to know whether a given piece of source code has or hasn’t been used to train a given deep learning model. On one side, it helps avoid data contamination problems that may exaggerate the performance of evaluated models. Conversely, it facilitates copyright protection by identifying private or protected code leveraged for model training without permission. To this end, automated approaches have been proposed for the detection, known as data contamination detection. Such approaches often heavily rely on the confidence of the involved models, assuming that the models should be more confident in handling contaminated data than cleaned data. However, such approaches do not consider the nature of the given data item, i.e., how difficult it is to predict the given item. Consequently, difficult-to-predict contaminated data and easy-to-predict cleaned data are often misclassified. As an initial attempt to solve this problem, this paper presents a naturalness-based approach, called Natural-DaCoDe , for code-completion models to distinguish contaminated source code from cleaned ones. Natural-DaCoDe leverages code naturalness to quantitatively measure the difficulty of a given source code for code-completion models. It then trains a classifier to distinguish contaminated source code according to both code naturalness and the performance of the code-completion models on the given source code. We evaluate Natural-DaCoDe with two pre-trained large language models (e.g., ChatGPT and Claude ) and two code-completion models that we trained from scratch for detecting contamination data. Our evaluation results suggest that Natural-DaCoDe substantially outperformed the state-of-the-art approaches in detecting contaminated data, improving the average accuracy by 61.78%. We also evaluate Natural-DaCoDe with method name suggestion task, and it remains more accurate than the state-of-the-art approaches, improving the accuracy by 54.39%. Furthermore, Natural-DaCoDe was tested on a natural language text benchmark, significantly outperforming the state-of-the-art approaches by 22% . It may suggest that Natural-DaCoDe could be applied to various source code related tasks besides code complete.