Избавляемся от ошибок Segmentation fault из-за переполнения стека в С++

Развивая идею доверенного языка программирования я пришел к выводу, что за счет ограничений синтаксиса и создания соответствующих проверок в статическом анализаторе кода, можно защититься практически ото всех технических ошибок, кроме двух - контроль динамически выделяемой памяти и переполнения стека. Причем, если для подсчета ссылок в рантайме, решения существуют, то контроль переполнения стека невозможно сделать не только во время анализа исходного текста программы, но это практически невозможно и во время выполнения приложения! Ведь ошибка переполнение стека (stack overflow) - это всегда фатально, так как не существует способа поймать и обработать эту ошибку изнутри выполняемой программы, чтобы потом продолжить её выполнение как ни в чем не бывало. Существует ли хотя бы теоретическая возможность защититься от ошибок переполнения стека и сделать из нее обычную ошибку (исключение), которую можно поймать (обработать) в самом приложении, чтобы была возможность продолжить выполнение программы без боязни последующей ошибки сегментации (segmentation fault) или повреждения стека (stack smashing)?

https://habr.com/ru/articles/983394/

#переполнение_стека #segmentation #segmentation_fault #stack_overflow #ошибки #исключения_в_c++ #обработка_ошибок #надежное_программирование #сезон_ии_в_разработке

Избавляемся от ошибок Segmentation fault из-за переполнения стека в С++

Развивая идею доверенного языка программирования я пришел к выводу, что за счет ограничений синтаксиса и создания соответствующих проверок в статическом анализаторе кода, можно защититься практически...

Хабр

Meta Releases SAM 3 for Enhanced Segmentation

https://techlife.blog/posts/meta-releases-sam-3/

#AI #Meta #SAM #Segmentation

Meta Releases SAM 3 for Enhanced Segmentation

Meta's latest update to its Segment Anything Model (SAM) brings significant improvements in accuracy and robustness.

TechLife

Meta SAM 3 공개: 텍스트만으로 영상 속 객체 추적, 정확도 2배 향상

Meta SAM 3는 텍스트 프롬프트만으로 영상 속 객체를 검출·추적합니다. 기존 대비 2배 향상된 정확도와 오픈소스 공개로 창작·개발 도구의 새 가능성을 엽니다.

https://aisparkup.com/posts/6730

sentencex - by Wikimedia:

https://github.com/wikimedia/sentencex

A sentence segmentation library with wide language support optimized for speed and utility.

Written in #Rust.

Bindings are available for #Python, #NodeJS and #WASM

Might be useful for my #SpeechToText system! 👀

#NLP #TextProcessing #Segmentation #RustLang

GitHub - wikimedia/sentencex: A sentence segmentation library with wide language support optimized for speed and utility.

A sentence segmentation library with wide language support optimized for speed and utility. - wikimedia/sentencex

GitHub
Why the overcomplication? Keep your email marketing segmentation simple and to the point, and avoid these common pitfalls. Cook up a sound segmentation strategy with our help: https://mailreach.eu/email-marketing-blog/segmentation-avoiding-overcomplication/?utm_medium=social&utm_source=mastodon&utm_campaign=blog
#EmailMarketing
#Newsletter
#EmailSegmentation
#Segmentation
#Marketing
#ReachRascals

So how do you turn contours (see previous post) into surface geometry? Well you can create a "levelset image", it is a bit like a heat propagating positively outward and negatively inward from the contours. Instead of heat though you can use the distance to the contour. Next you can draw the isosurface at the level 0 to retrieve the surface.

If one segments stuff using voxel labels it can be hard to retrieve a smooth result. Here however the contours are smoothly interpolated curves and the isosurface is smoothly interpolated as well. Hence there is naturally less of a stepped-Lego issue :).

In the video below the image on the right is the wobbly STL that comes with the Visible Human project. I tried to do better based on the contours-level set approach. I can also control the surface mesh density as you can see in the middle image.

#opensource #Julialang #biomedicalengineering #finiteelementanalysis #biomechanics #segmentation

Open source projects used here:

https://github.com/COMODO-research/Imago.jl

https://github.com/COMODO-research/Comodo.jl

https://github.com/COMODO-research/Geogram.jl

Unsupervised Instance Segmentation with Superpixels

Instance segmentation is essential for numerous computer vision applications, including robotics, human-computer interaction, and autonomous driving. Currently, popular models bring impressive performance in instance segmentation by training with a large number of human annotations, which are costly to collect. For this reason, we present a new framework that efficiently and effectively segments objects without the need for human annotations. Firstly, a MultiCut algorithm is applied to self-supervised features for coarse mask segmentation. Then, a mask filter is employed to obtain high-quality coarse masks. To train the segmentation network, we compute a novel superpixel-guided mask loss, comprising hard loss and soft loss, with high-quality coarse masks and superpixels segmented from low-level image features. Lastly, a self-training process with a new adaptive loss is proposed to improve the quality of predicted masks. We conduct experiments on public datasets in instance segmentation and object detection to demonstrate the effectiveness of the proposed framework. The results show that the proposed framework outperforms previous state-of-the-art methods.

arXiv.org

Senior Digital Marketer (JB5615)
Fully Remote (South African based Only)
$1 800.00 to $2 200.00 a month CTC

#Email #Ecommerce #Klaviyo #Mailchimp #Figma #Canva #SocialMedia #ABTesting #Segmentation #Analytics

Apply: https://bit.ly/DigitalMarketerJB5615
@KontakRecruitment

📢Know someone? Retweet & Tag

Oct 16: Last, but not least: Dagmar Kainmüller (Helmholtz Imaging, @MDC_Berlin) on current machine learning models for image segmentation—including how to apply such models to large data.

Register for the series 👉 https://bit.ly/6-image-processing-tasks

@helmholtz
#imaging #Segmentation

OmniCloudMask is a Python library for state-of-the-art segmentation of clouds and cloud shadows in high- to moderate-resolution satellite imagery #segmentation

https://github.com/DPIRD-DMA/OmniCloudMask

GitHub - DPIRD-DMA/OmniCloudMask

Contribute to DPIRD-DMA/OmniCloudMask development by creating an account on GitHub.

GitHub