Optical aberrations significantly degrade image quality in microscopy, particularly when imaging deeper into samples. These aberrations arise from distortions in the optical wavefront and can be mathematically represented using Zernike polynomials. Existing methods often address only mild aberrations on limited sample types and modalities, typically treating the problem as a black-box mapping without leveraging the underlying optical physics of wavefront distortions. We propose ZRNet, a physics-informed framework that jointly performs Zernike coefficient prediction and optical image Restoration. We contribute a Zernike Graph module that explicitly models physical relationships between Zernike polynomials based on their azimuthal degrees—ensuring that learned corrections align with fundamental optical principles. To further enforce physical consistency between image restoration and Zernike prediction, we introduce a Frequency-Aware Alignment (FAA) loss, which better aligns Zernike coefficient prediction and image features in the Fourier domain. Extensive experiments on CytoImageNet demonstrates that our approach achieves state-of-the-art performance in both image restoration and Zernike coefficient prediction across diverse microscopy modalities and biological samples with complex, large-amplitude aberrations. We further validate on experimental PSF data from a physical microscope and demonstrate robustness to realistic sensor noise, confirming generalisation beyond simulated conditions. Code is available at https://github.com/janetkok/ZRNet .

Crowd density estimation is a well-known computer vision task aimed at estimating the density distribution of people in an image. The primary challenge in this domain is the reliance on fine-grained location-level annotations — i.e., points placed on top of each individual — to train deep networks. Collecting such detailed annotations is both tedious, time-consuming, and poses a significant barrier to scalability for real-world applications. To alleviate this burden, we present Count2Density : a novel pipeline designed to predict meaningful density maps containing quantitative spatial information using only count-level annotations (i.e., the total number of people) during training. To achieve this, Count2Density generates pseudo-density maps leveraging past predictions stored in a Historical Map Bank, thereby reducing confirmation bias. This bank is initialised using an unsupervised saliency estimator to provide an initial spatial prior and is iteratively updated with an Exponential Moving Average of predicted density maps. These pseudo-density maps are obtained by sampling locations from estimated crowd areas using a hypergeometric distribution, with the number of samplings determined by the count-level annotations. To further enhance the spatial awareness of the model and promote robust feature learning, we add a self-supervised contrastive spatial regulariser to encourage similar feature representations within crowded regions while maximising dissimilarity with background regions. Experimental results demonstrate that our approach significantly outperforms cross-domain adaptation methods and achieves better results than recent state-of-the-art approaches in semi-supervised settings across several datasets. Additional analyses validate the effectiveness of each individual component of our pipeline, including the self-supervised contrastive regulariser, confirming the ability of Count2Density to effectively retrieve spatial information from count-level annotations and enabling accurate subregion counting.

Zero-Shot object counting aims to count instances of arbitrary object categories specified by text descriptions. Existing methods typically rely on vision-language models like CLIP, but often exhibit limited sensitivity to text prompts. We present T21 Count, a diffusion-based framework that lever-ages rich prior knowledge and fine-grained visual understanding from pretrained diffusion models. While one-step demising ensures efficiency, it leads to weakened text sensitivity. To address this challenge, we propose a Hierarchical Semantic Correction Module that progressively refines text-image feature alignment, and a Representational Regional Coherence Loss that provides reliable supervision signals by leveraging the cross-attention maps extracted from the demising U-Net. Furthermore, we observe that current benchmarks mainly focus on majority objects in images, potentially masking models' text sensitivity. To address this, we contribute a challenging re-annotated subset of FSC147 for better evaluation of text-guided counting ability. Extensive experiments demonstrate that our method achieves superior performance across different benchmarks. Code is available at https://github.com/chal5yq/T2lCount.

Saliency ranking detection (SRD) has emerged as a challenging task in computer vision, aiming not only to identify salient objects within images but also to rank them based on their degree of saliency. Existing SRD datasets have been created primarily using mouse-trajectory data, which inadequately captures the intricacies of human visual perception. Addressing this gap, this paper introduces the first large-scale SRD dataset, SIFR, constructed using genuine human fixation data, thereby aligning more closely with real visual perceptual processes. To establish a baseline for this dataset, we propose QAGNet, a novel model that leverages salient instance query features from a transformer detector within a tri-tiered nested graph. Through extensive experiments, we demonstrate that our approach outperforms existing state-of-the-art methods across two widely used SRD datasets and our newly proposed dataset. Code and dataset are available at https://github.com/EricDengbowen/QAGNet.

Salient object detection plays an important role in many downstream tasks. However, complex real-world scenes with varying scales and numbers of salient objects still pose a challenge. In this paper, we directly address the problem of detecting multiple salient objects across complex scenes. We propose a network architecture incorporating non-local feature information in both the spatial and channel spaces, capturing the long-range dependencies between separate objects. Traditional bottom-up and non-local features are combined with edge features within a feature fusion gate that progressively refines the salient object prediction in the decoder. We show that our approach accurately locates multiple salient regions even in complex scenarios. To demonstrate the efficacy of our approach to the multiple salient objects problem, we curate a new dataset containing only multiple salient objects. Our experiments demonstrate the proposed method presents state-of-the-art results on five widely used datasets without any pre-processing and post-processing. We obtain a further performance improvement against competing techniques on our multi-objects dataset. The dataset and source code are available at: https://github.com/EricDengbowen/DSLRDNet.

Self-supervised learning has emerged as a candidate approach to learn semantic visual features from unlabeled video data. In self-supervised learning, intrinsic correspondences between data points are used to define a proxy task that forces the model to learn semantic representations. Most existing proxy tasks applied to video data exploit only either intra-modal (e.g. temporal) or cross-modal (e.g. audio-visual) correspondences separately. In theory, jointly learning both these correspondences may result in richer visual features; but, as we show in this work, doing so is non-trivial in practice. To address this problem, we introduce `Audio-Visual Permutative Predictive Coding' (AV-PPC), a multi-task learning framework designed to fully leverage the temporal and cross-modal correspondences as natural supervision signals. In AV-PPC, the model is trained to simultaneously learn multiple intra- and cross-modal predictive coding sub-tasks. By using visual speech recognition (lip-reading) as the downstream evaluation task, we show that our proposed proxy task can learn higher quality visual features than existing proxy tasks. We also show that AV-PPC visual features are highly data-efficient. Without further finetuning, AV-PPC visual encoder achieves 80.30% spoken word classification rate on the LRW dataset, performing on par with directly supervised visual encoders that are learned from large amounts of labeled data.

Plant phenotyping is the quantitative description of a plant's physiological, biochemical, and anatomical status which can be used in trait selection and helps to provide mechanisms to link underlying genetics with yield. Here, an active vision- based pipeline is presented which aims to contribute to reducing the bottleneck associated with phenotyping of architectural traits. The pipeline provides a fully automated response to photometric data acquisition and the recovery of three-dimensional (3D) models of plants without the dependency of botanical expertise, whilst ensuring a non-intrusive and non-destructive approach. Access to complete and accurate 3D models of plants supports computation of a wide variety of structural measurements. An Active Vision Cell (AVC) consisting of a camera-mounted robot arm plus combined software interface and a novel surface reconstruction algorithm is proposed. This pipeline provides a robust, flexible, and accurate method for automating the 3D reconstruction of plants. The reconstruction algorithm can reduce noise and provides a promising and extendable framework for high throughput phenotyping, improving current state-of-the-art methods. Furthermore, the pipeline can be applied to any plant species or form due to the application of an active vision framework combined with the automatic selection of key parameters for surface reconstruction.

We address the complex problem of reliably segmenting root structure from soil in X-ray Computed Tomography (CT) images. We utilise a deep learning approach, and propose a state-of-the-art multi-resolution architecture based on encoderdecoders. While previous work in encoder-decoders implies the use of multiple resolutions simply by downsampling and upsampling images, we make this process explicit, with branches of the network tasked separately with obtaining local high-resolution segmentation, and wider low-resolution contextual information. The complete network is a memory efficient implementation that is still able to resolve small root detail in large volumetric images. We compare against a number of different encoder-decoder based architectures from the literature, as well as a popular existing image analysis tool designed for root CT segmentation. We show qualitatively and quantitatively that a multi-resolution approach offers substantial accuracy improvements over a both a small receptive field size in a deep network, or a larger receptive field in a shallower network. We then further improve performance using an incremental learning approach, in which failures in the original network are used to generate harder negative training examples. Our proposed method requires no user interaction, is fully automatic, and identifies large and fine root material throughout the whole volume.

The emerging discipline of plant phenomics aims to measure key plant characteristics, or traits, though as yet the set of plant traits that should be measured by automated systems is not well defined. Methods capable of recovering generic representations of the 3D structure of plant shoots from images would provide a key technology underpinning quantification of a wide range of current and future physiological and morphological traits. We present a fully automatic approach to image-based 3D plant reconstruction which represents plants as series of small planar sections that together model the complex architecture of leaf surfaces. The initial boundary of each leaf patch is refined using a level set method, optimising the model based on image information, curvature constraints and the position of neighbouring surfaces. The reconstruction process makes few assumptions about the nature of the plant material being reconstructed. As such it is applicable to a wide variety of plant species and topologies, and can be extended to canopy-scale imaging. We demonstrate the effectiveness of our approach on real images of wheat and rice plants, an artificial plant with challenging architecture, as well as a novel virtual dataset that allows us to compute distance measures of reconstruction accuracy. We also illustrate the method’s potential to support the identification of individual leaves, and so the phenotyping of plant shoots, using a spectral clustering approach.