共找到 20 条结果
Commercial plant phenotyping systems using fixed cameras cannot perceive many plant details due to leaf occlusion. In this paper, we present Botany-Bot, a system for building detailed "annotated digital twins" of living plants using two stereo cameras, a digital turntable inside a lightbox, an industrial robot arm, and 3D segmentated Gaussian Splat models. We also present robot algorithms for manipulating leaves to take high-resolution indexable images of occluded details such as stem buds and the underside/topside of leaves. Results from experiments suggest that Botany-Bot can segment leaves with 90.8% accuracy, detect leaves with 86.2% accuracy, lift/push leaves with 77.9% accuracy, and take detailed overside/underside images with 77.3% accuracy. Code, videos, and datasets are available at https://berkeleyautomation.github.io/Botany-Bot/.
Foundation models have demonstrated a remarkable ability to learn rich, transferable representations across diverse modalities such as images, text, and audio. In modern machine learning pipelines, these representations often replace raw data as the primary input for downstream tasks. In this paper, we address the challenge of adapting a pre-trained foundation model to inject domain-specific knowledge, without retraining from scratch or incurring significant computational costs. To this end, we introduce BotaCLIP, a lightweight multimodal contrastive framework that adapts a pre-trained Earth Observation foundation model (DOFA) by aligning high-resolution aerial imagery with botanical relevés. Unlike generic embeddings, BotaCLIP internalizes ecological structure through contrastive learning with a regularization strategy that mitigates catastrophic forgetting. Once trained, the resulting embeddings serve as transferable representations for downstream predictors. Motivated by real-world applications in biodiversity modeling, we evaluated BotaCLIP representations in three ecological tasks: plant presence prediction, butterfly occurrence modeling, and soil trophic group abundance estim
According to the European Union's Habitat Directive, habitat monitoring plays a critical role in response to the escalating problems posed by biodiversity loss and environmental degradation. Scree habitats, hosting unique and often endangered species, face severe threats from climate change due to their high-altitude nature. Traditionally, their monitoring has required highly skilled scientists to conduct extensive fieldwork in remote, potentially hazardous locations, making the process resource-intensive and time-consuming. This paper presents a novel approach for scree habitat monitoring using a legged robot to assist botanists in data collection and species identification. Specifically, we deployed the ANYmal C robot in the Italian Alpine bio-region in two field campaigns spanning two years and leveraged deep learning to detect and classify key plant species of interest. Our results demonstrate that agile legged robots can navigate challenging terrains and increase the frequency and efficiency of scree monitoring. When paired with traditional phytosociological surveys performed by botanists, this robotics-assisted protocol not only streamlines field operations but also enhances
Over two million containers crossed the docks at Sydney's Port Botany in 2011/12; a figure that is forecast increase more than threefold by the end of the next decade. To cope with such large growth in volumes the NSW Government plans to double rail mode share at the port by the year 2020. Conventional wisdom from industry and the media says that existing infrastructure cannot handle such volumes. In this paper we use a combination of data analytics and simulation to examine operations at the port and evaluate the efficacy of current infrastructure to handle projected growth in volumes. Contrary to conventional wisdom we find that current rail resources appear distinctly under-utilised. Moreover: (i) the peak rail capacity of Port Botany is 1.78 million TEU per annum; over six times higher than 2011/12 rail volumes; (ii) there are no infrastructural impediments to the achievement of peak rail capacity; (iii) operational changes, not infrastructural investment, are the key to unlocking the potential of the port; (iv) Port Botany is well positioned to handle projected increases in container volumes over the next decade and beyond, including the 28% rail mode share target established
The geography and botany of smooth/symplectic nonspin 4-manifolds with abelian fundamental group are addressed.
We give a classification of iwip outer automorphisms of the free group, by discussing the properties of their attracting and repelling trees.
This note explores two questions: (1) Which bigraded groups arise as the knot Floer homology of a knot in the three-sphere? (2) Given a knot, how many distinct knots share its Floer homology? Regarding the first, we show there exist bigraded groups satisfying all previously known constraints of knot Floer homology which do not arise as the invariant of a knot. This leads to a new constraint for knots admitting lens space surgeries, as well as a proof that the rank of knot Floer homology detects the trefoil knot. For the second, we show that any non-trivial band sum of two unknots gives rise to an infinite family of distinct knots with isomorphic knot Floer homology. We also prove that the fibered knot with identity monodromy is strongly detected by its knot Floer homology, implying that Floer homology solves the word problem for mapping class groups of surfaces with non-empty boundary. Finally, we survey some conjectures and questions and, based on the results described above, formulate some new ones.
Detecting heavy metal pollution in soils and seaports is vital for regional environmental monitoring. The Pollution Load Index (PLI), an international standard, is commonly used to assess heavy metal containment. However, the conventional PLI assessment involves laborious procedures and data analysis of sediment samples. To address this challenge, we propose a deep-learning-based model that simplifies the heavy metal assessment process. Our model tackles the issue of data scarcity in the water-sediment domain, which is traditionally plagued by challenges in data collection and varying standards across nations. By leveraging transfer learning, we develop an accurate quantitative assessment method for predicting PLI. Our approach allows the transfer of learned features across domains with different sets of features. We evaluate our model using data from six major ports in New South Wales, Australia: Port Yamba, Port Newcastle, Port Jackson, Port Botany, Port Kembla, and Port Eden. The results demonstrate significantly lower Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) of approximately 0.5 and 0.03, respectively, compared to other models. Our model performance i
Li-Xie-Zhang classified instanton Floer minimal knots in balanced sutured manifolds subject to a condition on the fundamental group. In this paper, we give a similar classification in the Heegaard Floer homology setting. Since our classifications agree when they are both applicable, this provides further evidence for the conjecture of Kronheimer-Mrowka that instanton Floer homology and Heegaard Floer homology are isomorphic. We also study link Floer homology botany question in $S^1\times S^2$, showing that link Floer homology detects spherical braid closures among homologically nontrivial links.
The LifeCLEFs plant identification task provides a testbed for a system-oriented evaluation of plant identification about 500 species trees and herbaceous plants. Seven types of image content are considered: scan and scan-like pictures of leaf, and 6 kinds of detailed views with unconstrained conditions, directly photographed on the plant: flower, fruit, stem & bark, branch, leaf and entire view. The main originality of this data is that it was specifically built through a citizen sciences initiative conducted by Tela Botanica, a French social network of amateur and expert botanists. This makes the task closer to the conditions of a real-world application. This overview presents more precisely the resources and assessments of task, summarizes the retrieval approaches employed by the participating groups, and provides an analysis of the main evaluation results. With a total of ten groups from six countries and with a total of twenty seven submitted runs, involving distinct and original methods, this fourth year task confirms Image & Multimedia Retrieval community interest for biodiversity and botany, and highlights further challenging studies in plant identification.
Plant roots typically exhibit a highly complex and dense architecture, incorporating numerous slender lateral roots and branches, which significantly hinders the precise capture and modeling of the entire root system. Additionally, roots often lack sufficient texture and color information, making it difficult to identify and track root traits using visual methods. Previous research on roots has been largely confined to 2D studies; however, exploring the 3D architecture of roots is crucial in botany. Since roots grow in real 3D space, 3D phenotypic information is more critical for studying genetic traits and their impact on root development. We have introduced a 3D root skeleton extraction method that efficiently derives the 3D architecture of plant roots from a few images. This method includes the detection and matching of lateral roots, triangulation to extract the skeletal structure of lateral roots, and the integration of lateral and primary roots. We developed a highly complex root dataset and tested our method on it. The extracted 3D root skeletons showed considerable similarity to the ground truth, validating the effectiveness of the model. This method can play a significant
The 2017-th edition of the LifeCLEF plant identification challenge is an important milestone towards automated plant identification systems working at the scale of continental floras with 10.000 plant species living mainly in Europe and North America illustrated by a total of 1.1M images. Nowadays, such ambitious systems are enabled thanks to the conjunction of the dazzling recent progress in image classification with deep learning and several outstanding international initiatives, such as the Encyclopedia of Life (EOL), aggregating the visual knowledge on plant species coming from the main national botany institutes. However, despite all these efforts the majority of the plant species still remain without pictures or are poorly illustrated. Outside the institutional channels, a much larger number of plant pictures are available and spread on the web through botanist blogs, plant lovers web-pages, image hosting websites and on-line plant retailers. The LifeCLEF 2017 plant challenge presented in this paper aimed at evaluating to what extent a large noisy training dataset collected through the web and containing a lot of labelling errors can compete with a smaller but trusted trainin
We present new families of examples of non-simple prime Legendrian and transversal knots in tight Lens spaces, which demonstrate that the botany of Legendrians in Lens space is rich. In fact, there are more non-isotopic Legendrians that are topologically isotopic to the $n$-twist knot in a Lens space $L(α, β)$ than in $S^3$. We also include connect sum formulas for rational variants of classical invariants, $\mathrm{tb}_\mathbb{Q}$, $\mathrm{rot}_\mathbb{Q}$, and $\mathrm{sl}_\mathbb{Q}$, which indicate that prime knots are the right playground to look for exotic behaviour.
The conceptual landscape of convection has two simple gateways: optimal function and random form. Optimal convection adjusts toward a univariate ideal called neutrality. Convection form involves elements (parcels, bubbles, drafts) whose most parsimonious assumption is random. Between these gates lies a wilderness of realizable flow configurations. The only simple principle is natural selection by fitness, a scalar whose gradient is a local direction in an abstract configuration space. Random or high-entropy patterns occupy most of configuration space and occur spontaneously. With time, convection can discover less facile but more efficient (organized) configurations, by sequential selection. Here two data exercises explore that self-organization process, in shallow and deep moist convection. For shallow convection, causal network postulates are explored in a large set of cyclic-domain large-eddy simulations (LES; the Cloud Botany set). When an evolutionary pathway (mainly layer deepening in these simulations) leads to precipitation, mesoscale patterns blossom rapidly. For deep convection, expanding rings of conditional cell probability around prior cells are estimated from satellit
Let $α$ be a map from the set of all knot types ${\mathcal K}$ to a set $X$. Let $β$ be a map from ${\mathcal K}$ to a set $Y$. We define the relation between $α$ and $β$ to be the image of a map $(α,β)$ from ${\mathcal K}$ to $X\times Y$ sending an element $K$ of ${\mathcal K}$ to $(α(K),β(K))$. We determine the relations between $α$ and $β$ for certain $α$ and $β$ such as crossing number, unknotting number, bridge number, braid index, genus and canonical genus. This is a study of geography problem in knot theory.
Phyllotactic patterns, i.e. regular arrangements of leaves or seeds around a plant stem, are fascinating examples of complex structures encountered in Nature. In botany, their symmetries develop when a new primordium periodically grows in the largest gap left between the previous primordium and the apex. Experiments using ferrofluid droplets have also shown that phyllotactic patterns can spontaneously form when identical elements repulsing each other are periodically released at a given distance from an injection center and are advected radially at a constant speed. A central issue in phyllotaxis is to understand whether other self-organized mechanisms can generate such patterns. Here, we show that phyllotactic structures also develop in the large class of spatial symmetry-breaking systems giving spotted patterns with an intrinsic wavelength, in the case of radial growth. We evidence this experimentally on chemical precipitation patterns, and numerically on two different models describing reaction-driven phase transitions and spatial Turing patterns, respectively. A generalized method for the construction of this new family of phyllotactic structures is presented, which paves the w
The domain of Botany is rich with metaphorical terms. Those terms play an important role in the description and identification of flowers and plants. However, the identification of such terms in discourse is an arduous task. This leads in some cases to committing errors during translation processes and lexicographic tasks. The process is even more challenging when it comes to machine translation, both in the cases of single-word terms and multi-word terms. One of the recent concerns of Natural Language Processing (NLP) applications and Machine Translation (MT) technologies is the automatic identification of metaphor-based words in discourse through Deep Learning (DL). In this study, we seek to fill this gap through the use of thirteen popular transformer based models, as well as ChatGPT, and we show that discriminative models perform better than GPT-3.5 model with our best performer reporting 92.2349% F1 score in metaphoric flower and plant names identification task.
The detection of leaf diseases in plants generally involves visual observation of patterns appearing on the leaf surface. However, there are many diseases that are distinguished based on very subtle changes in these visually observable patterns. This paper attempts to identify plant leaf diseases using image processing techniques. The focus of this study is on the detection of citrus leaf canker disease. Canker is a bacterial infection of leaves. Symptoms of citrus cankers include brown spots on the leaves, often with a watery or oily appearance. The spots (called lesions in botany) are usually yellow. It is surrounded by a halo of the leaves and is found on both the top and bottom of the leaf. This paper describes various methods that have been used to detect citrus leaf canker disease. The methods used are histogram comparison and k-means clustering. Using these methods, citrus canker development was detected based on histograms generated based on leaf patterns. The results thus obtained can be used, after consultation with experts in the field of agriculture, to identify suitable treatments for the processes used.
Elliptic Fourier analysis (EFA) is a powerful tool for shape analysis, which is often employed in geometric morphometrics. However, the normalization of elliptic Fourier descriptors has persistently posed challenges in obtaining unique results in basic contour transformations, requiring extensive manual alignment. Additionally, contemporary contour/outline extraction methods often struggle to handle complex digital images. Here, we reformulated the procedure of EFDs calculation to improve computational efficiency and introduced a novel approach for EFD normalization, termed true EFD normalization, which remains invariant under all basic contour transformations. These improvements are crucial for processing large sets of contour curves collected from different platforms with varying transformations. Based on these improvements, we developed ElliShape, a user-friendly software. Particularly, the improved contour/outline extraction employs an interactive approach that combines automatic contour generation for efficiency with manual correction for essential modifications and refinements. We evaluated ElliShape's stability, robustness, and ease of use by comparing it with existing softw