Various levels of background and noise were introduced into a total of 692 190 simulated patterns to assess their effects. Dataset 1 was devoid of background but included Gaussian noise with a standard deviation ( σ ) of 0.25. Dataset 2 featured a 3% background ratio along with Gaussian noise ( σ = 0.25). Datasets 3 and 4 consisted of Gaussian noise with standard deviations of σ = 1 and σ = 3, respectively, without any background. Table S2 in Supplementary Note 2 of the supporting information shows that CPICANN's accuracy in single-phase identification without elemental information was 86.98% for the 3% background mixture, 86.35% for the noise mixture of σ = 1 and 84.30% for the noise mixture of σ = 3. Evidently, high levels of background or noise adversely affect CPICANN's performance. All datasets and corresponding pre-trained models are publicly accessible. Specific background-stripping algorithms and smoothing techniques could alleviate these challenges during real-world analysis. Nevertheless, while striving for high-throughput autonomous characterization, some degree of accuracy may need to be compromised. However, subsequent discussions in this work will focus on dataset 1, which presents minimal noise interference, to concentrate on the phase-identification challenge.
( )–( ) Simulated X-ray powder diffraction patterns of PbSO crystal under a Cu anode. ( ) An ideal crystal; ( ) an average grain size of 3 nm; ( ) an orientation factor of 0.3, a thermal-vibration derivation of 0.2 and a zero shift of 1.2°; and ( ) with background intensity. |
In addition to the single-phase patterns, binary-phase XRD patterns were generated by leveraging the 692 190 simulated single-phase patterns. This was achieved by blending two patterns with the formula py 1 + (1 − p ) y 2 , where y 1 and y 2 represent selected single-phase patterns and p denotes the mixing ratio, ranging from 0.2 to 0.8. This blending approach assumes that no reaction occurs between the phases. A pre-process on powder XRD patterns is carried out by selecting 4500 points from each XRD pattern within the 2 θ range from 10 to 80° with a step size of 0.015° so that the corresponding intensities are expressed as a 4500 × 1-dimensional vector, which is hence standardized. The intensity for each specific diffraction angle is assigned based on the experimental patterns provided. In this study, we identify the nearest diffraction angle on the experimental pattern and allocate the corresponding intensity to the 4500-dimensional input vector. In cases where two angles are equally close to the matched angle, we select the one with the higher intensity.
The architecture of CPICANN. In each of the one-dimensional convolution layers, × 1 conv., and /2 denote the kernel size , the channel number and a stride of 2, respectively. In the max-pooling layers, /2 also indicates a stride of 2. Residual connection is indicated by solid lines. The convoluted information is fed into six eight-head self-attention blocks, which scores the input XRD pattern against the 23 073 single-phase patterns. |
CPICANN incorporates elemental information by using an element filter like that used in JADE . During the inference process with elemental information, the filter is applied on the model output and categorizes all elements in the periodic table into three groups: A `included elements', B `possible elements' and C `excluded elements', viz . all elements are A ∪ B ∪ C . In the XRD pattern identification, the included elements must appear simultaneously with variation in their individual concentrations, the possible elements can possibly appear individually and the excluded elements cannot appear at all. For example, if Fe is the included element and S and O are possible elements, the XRD patterns for crystals Fe, FeS, Fe 2 O 3 , Fe 3 O 4 , Fe 2 (SO 4 ) 3 , etc . form a set S , much smaller than the whole set, and an analyzed XRD pattern will be matched with those in set S . But, the XRD pattern for crystal FeCl 2 , for example, does not belong to the set S because Cl is one of the excluded elements.
( ) The data distribution in the seven crystal systems for both the training and testing datasets. ( ) The performances of CPICANN and Task-Macro in on the single-phase identification in each of the seven crystal systems. ( ) The performance accuracy versus random sample amounts of CPICANN and on the single-phase identification with elemental information, where the accuracy is averaged over the seven crystal systems. ( ) The performances of CPICANN and on the single-phase identification over 1000 random XRD patterns in each of the seven crystal systems. |
The probability of the right two phases in the recommended phases as a function of the number of recommended phases with and without elemental information. |
The probability of the right phase in the phases recommended by CPICANN and , where the dark and light colors mark recommended one and three phases, respectively. |
The present work develops a novel network, CPICANN, for crystal phase identification on whole X-ray powder diffraction patterns, utilizing a convolutional self-attention mechanism. CPICANN can automate and integrate the XRD patterns into a unified attention-matching strategy. The performance and effectiveness of CPICANN are extremely powerful, as shown here by the single-phase and bi-phase identifications with and without elemental information on simulated XRD patterns, and the single-phase identification on experimental XRD patterns. Elemental information is initially provided manually in the conventional identification approach. In contrast, CPICANN employs elemental information afterwards, applying it only to those highly potential crystal phases selected based on their attention probability scores from the examined whole XRD pattern. This merit of CPICANN minimizes potential errors in the elemental information provided.
The success of CPICANN in phase identification represents a significant advancement in materials informatics, providing a more efficient and accurate method for automatic phase identification and rapid screening in complex material crystal structures. In future work, we will integrate CPICANN with the XRD refinement software WPEM to develop an AI-driven XRD analyzer.
The model described in the present work was implemented in Python. Source codes are available at https://www.github.com/WPEM/CPICANN .
Supporting information. DOI: https://doi.org/10.1107/S2052252524005323/fc5077sup1.pdf
‡ These authors contributed equally to this work and should be considered as co-first authors.
Thanks to the International Centre for Diffraction Data for providing a trial version of JADE Pro 8.9 and PDF-5+ 2024 to Mr Cao Bin. We are deeply grateful for the invaluable guidance and insights provided by Professor Qian Quan from Shanghai University, which have greatly enriched this study.
The authors declare no competing interests.
All data studied in this article are available at https://www.github.com/WPEM/CPICANN .
The following funding is acknowledged: Shanghai Pujiang Program (Grant No. 23PJ1403500) and Guangzhou-HKUST(GZ) Joint Funding Program (No. 2023A03J0003 and No. 2023A03J0103).
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence , which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.
IMAGES
VIDEO
COMMENTS
Top 4 Essay Attention Getters. The top four types of attention-grabbing openings are: Asking the reader a question. Telling a story. Telling a joke. Making a comparison. These four attention getters can be used interchangeably in pretty much any type of essay.
The introductory paragraph of a research paper needs to grab the reader and make her want to continue reading. Through the use of quotes, questions, statistics or even relevant words, you can pull the reader in and get her excited about your topic.
Hooks and Attention GrabbersH. oks and Attention Grabbers The first sentence of your introduction is the first chance a writer has to capture. he attention of the reader. Some people call this a "hook" because it captures a reader's attention with interesting statements and ideas just like a fisherman will use a shiny lure to g.
The most compelling introductory lines include posing a question, storytelling, sharing fascinating statistics, and rhetorical generalizations. Below, we list examples of attention-getters for essays to help you get on the right foot. Whatever method you choose, you can't go wrong. Hooks go interchangeably in almost any written piece.
Learn about hooks for essays, which are commonly referred to as attention getters. Explore essay hook examples, tips for writing, and types of attention getters. Updated: 11/21/2023
Here are 7 writing hooks that make readers want to find out what you will say in the rest of your essay. Interesting Question Hook. Strong Statement/Declaration Hook. Fact/Statistic Hook. Metaphor/ Simile Hook. Story Hook. Description Hook. Quotation Hook. 1.
7 Tips for Writing an Attention-Grabbing Hook. How do you get a reader interested in what you have to say? One technique is to use a great hook—an opening so exciting that it convinces a reader that your story is worth reading. How do you get a reader interested in what you have to say? One technique is to use a great hook—an opening so ...
1. Tell a personal anecdote that relates to your topic. Pick an anecdote that will introduce your topic to the reader through setting, scene, and detail. Guide the reader through the story so they are drawn in. Try to keep the anecdote short and to the point, around two to four lines at the most.
When writing an essay, it's important to get the reader's attention from the very beginning. You want to make sure you immediately draw in the reader so that they are motivated to read the rest of the piece. This can be done by creating a strong "attention getter" in the beginning of the article.
Examples of Attention Getters in Different Formats Essays and Research Papers Thesis statements. Thesis statements are an effective way to grab reader's attention and focus the essay topic. A well-crafted thesis statement can be used to introduce a topic, provide insight into the main points of an essay, and act as a road map for readers so ...
A hook is the first sentence of the introductory paragraph of a paper or essay. The hook differs from a thesis statement in that the purpose of the hook is to get the reader's attention and ...
By Christopher Cascio. Attention grabbers are techniques you use at the very beginning of an essay as a means to hook your readers' attention and get them interested in your topic. You can use one of several techniques, such as a surprising statistic, a generalization or even a story. However, no matter which method you use, you need to make ...
The Neuron is a daily AI newsletter that tracks the latest AI trends and tools you need to know. Join 400,000+ professionals from top companies like Microsoft, Apple, Salesforce and more. 100% FREE. Having a strong opening is essential for any piece of writing. It is the attention getter that hooks the reader in and makes them want to keep reading.
An attention-getter is the device a speaker uses at the beginning of a speech to capture an audience's interest and make them interested in the speech's topic. Typically, there are four things to consider in choosing a specific attention-getting device: Appropriateness or relevance to audience. Purpose of speech. Topic.
Tell a Joke. 4. Use Comparisons. Attention Grabbers for a Persuasive or Argumentative Papers. Attention Grabbers for a Descriptive Paper. Attention Grabbers for Contrast Papers. Draw people into reading your essay (for example informative essay or 500 word essay) by making your introduction interesting and intriguing with the first sentence it ...
6. a Good Story. 7. Quotes. To Wrap Up. The first few seconds of your speech will determine whether or not your audience will give you their attention. This is why you should begin with attention-grabbing openings. Attention-getter examples include stories, quotes, and questions that persuades an audience to listen. Conveying a powerful message ...
An attention-getter is often referred to as the "hook" of the essay. A good hook makes the reader want to keep reading. It gets the reader interested in the topic. You can then set the stage with some background information on your topic. Once the reader is hooked, he or she will be warmed up for the thesis.
About This Quiz & Worksheet. Attention getters in academic essays are an important component of the essay, and this quiz and worksheet will help you to identify effective attention getters and how ...
Attention Getters for Different Purposes. As you already know, each hook type suits specific texts. In this next segment, we will discuss which attention getters to use with which assignments to get the best outcome. Attention Getter for an Argumentative Essay. Argumentative essays aim to engage the reader in a discussion.
Give the audience a reason to listen. Clear transition into the body of the speech. Table of Contents. 12 Attention Getters for Speeches. Ask a Rhetorical Question. Make a Bold Statement. State the importance. Use Humor. Shocking Statistics or Facts.
Different Content Types Where Our Attention Getter Generator Comes in Handy. Not every kind of document will be fitting for an introductory hook. For instance, the tool may not be a perfect option for works of fiction. Nevertheless, here are some top content types where you'd definitely find the attention getter generator super useful. Essays
3. An Intriguing Video. There's something captivating about the combination of imagery, animations, and audio in a video that captures attention. Here are a few examples of using compelling video footage. View example. View example. View example. 4. A Realistic Scenario.
Attention getters are a great classroom management tool! Whether students are busy working in groups, on class projects, or are just plain chatty, these attention getters are not only fun, but are an effective way to get your class re-focused, so you can give instructions. The ideas listed are interactive. The teacher says the first phrase, […]
The development of CPICANN, a novel convolutional self-attention neural network, represents a groundbreaking approach in materials informatics. By leveraging the convolutional self-attention mechanism, CPICANN automates and significantly enhances the efficiency of crystal phase identification from whole X-ray powder diffraction patterns, marking a substantial advancement over traditional time ...
First Opinion readers respond to essays about long Covid, FDA guidance on diversity in clinical trials, and more. ... To secure the U.S.'s biotech supply chain, pay attention to domestic ...