https://blog.bj-yan.top/en/tags/deep-learning/ Recent content in Deep-Learning on Bingjie's Blog https://blog.bj-yan.top/favicon.png https://blog.bj-yan.top/favicon.png Hugo -- gohugo.io en-US Wed, 09 Aug 2023 00:54:19 +0800 https://blog.bj-yan.top/en/p/blog-efficiency-analysis-of-distributed-machine-learning/ Wed, 09 Aug 2023 00:54:19 +0800 https://blog.bj-yan.top/en/p/blog-efficiency-analysis-of-distributed-machine-learning/ Analyzing the stochastic runtime of synchronous, asynchronous, and K-worker SGD, and discussing the time to reach target accuracy and parallel speedup limits in conjunction with convergence bounds. https://blog.bj-yan.top/en/p/blog-convergence-analysis-in-deep-learning-part-5/ Tue, 08 Aug 2023 15:04:00 +0800 https://blog.bj-yan.top/en/p/blog-convergence-analysis-in-deep-learning-part-5/ Deriving non-convex stationary point bounds and strongly convex error bounds under full participation from FedAvg's local updates, distinguishing gradient noise, client drift, and sampling error, while discussing partial participation and heterogeneity corrections. https://blog.bj-yan.top/en/p/blog-convergence-analysis-in-deep-learning-part-4/ Mon, 07 Aug 2023 15:03:00 +0800 https://blog.bj-yan.top/en/p/blog-convergence-analysis-in-deep-learning-part-4/ Deriving strong convexity convergence bounds for synchronous, asynchronous, and K-Async SGD, clarifying assumptions of independent sampling, conditional unbiasedness, gradient staleness, and freshness probability. https://blog.bj-yan.top/en/p/blog-privacy-leakage-in-deep-learning/ Fri, 28 Jul 2023 09:54:50 +0800 https://blog.bj-yan.top/en/p/blog-privacy-leakage-in-deep-learning/ Reviewing membership inference, attribute inference, model inversion, and training data extraction; deriving the recoverability of single-sample gradients; and discussing the applicability boundaries of federated learning, risk assessment, and various privacy-preserving techniques. https://blog.bj-yan.top/en/p/blog-convergence-analysis-in-deep-learning-part-3/ Tue, 25 Jul 2023 15:03:00 +0800 https://blog.bj-yan.top/en/p/blog-convergence-analysis-in-deep-learning-part-3/ Deriving strong convex error bounds and non-convex stationary point bounds for mini-batch SGD under assumptions of conditionally unbiased gradients and bounded variance, explaining the impact of step size and noise. https://blog.bj-yan.top/en/p/blog-convergence-analysis-in-deep-learning-part-2/ Mon, 24 Jul 2023 15:02:00 +0800 https://blog.bj-yan.top/en/p/blog-convergence-analysis-in-deep-learning-part-2/ Derive first-order method convergence bounds under smooth/non-smooth and convex/non-convex conditions, clarifying step sizes, output points, and iteration complexity. https://blog.bj-yan.top/en/p/blog-convergence-analysis-in-deep-learning-part-1/ Sun, 23 Jul 2023 15:01:00 +0800 https://blog.bj-yan.top/en/p/blog-convergence-analysis-in-deep-learning-part-1/ Unify definitions and assumptions regarding convexity, strong convexity, smoothness, and stochastic gradients, and derive inequalities commonly used in convergence analysis. https://blog.bj-yan.top/en/p/blog-will-ai-replace-the-artists/ Mon, 12 Sep 2022 12:01:09 +0800 https://blog.bj-yan.top/en/p/blog-will-ai-replace-the-artists/ Combining the development of diffusion models and AI art, this article discusses which tasks in artistic creation might be automated, along with the capabilities, limitations, and the creator's technical role in image generation. https://blog.bj-yan.top/en/p/blog-diffusion-model-trial/ Sun, 11 Sep 2022 23:03:04 +0800 https://blog.bj-yan.top/en/p/blog-diffusion-model-trial/ Reviewing the basic concepts and development history of diffusion models, and recording trial experiences and generation results of image generation models such as Stable Diffusion and DALL·E. https://blog.bj-yan.top/en/p/blog-hcaptcha-seaplane/ Sun, 01 May 2022 23:35:46 +0800 https://blog.bj-yan.top/en/p/blog-hcaptcha-seaplane/ Records the training and testing practices for the hCaptcha seaplane recognition task, discussing sample annotation, overfitting issues, and the organization of the model training workflow. https://blog.bj-yan.top/en/p/blog-where-is-the-next-generation-of-captcha/ Sun, 01 May 2022 01:53:59 +0800 https://blog.bj-yan.top/en/p/blog-where-is-the-next-generation-of-captcha/ Reviewing the evolution of CAPTCHA from character recognition to image recognition, discussing human-machine confrontation, user experience, and potential directions for the next generation of CAPTCHA in the context of hCaptcha practices. https://blog.bj-yan.top/en/p/blog-hcaptcha-elephant-drawn-with-leaves/ Sun, 01 May 2022 01:53:15 +0800 https://blog.bj-yan.top/en/p/blog-hcaptcha-elephant-drawn-with-leaves/ Analyze the data and visual characteristics of the hCaptcha leaf elephant image recognition task, discussing sample classification, model training, and identification challenges caused by similar images. https://blog.bj-yan.top/en/p/blog-hcaptcha-sky-left-airplane/ Thu, 03 Mar 2022 23:34:48 +0800 https://blog.bj-yan.top/en/p/blog-hcaptcha-sky-left-airplane/ Analyze the hCaptcha task of identifying left-flying airplanes, introducing an implementation approach that uses background complexity, image color, and directional features to determine if an airplane is in the sky. https://blog.bj-yan.top/en/p/blog-hcaptcha-vertical-river/ Wed, 02 Mar 2022 10:32:15 +0800 https://blog.bj-yan.top/en/p/blog-hcaptcha-vertical-river/ Analyze the hCaptcha vertical river image task, document observations of image features, recognition methods, and solution iterations, and discuss the impact of generated images on CAPTCHA recognition. https://blog.bj-yan.top/en/p/blog-deep-learning-for-pixel-art/ Tue, 15 Feb 2022 01:19:35 +0800 https://blog.bj-yan.top/en/p/blog-deep-learning-for-pixel-art/ Organizing the stylistic features, galleries, datasets, and drawing tools of pixel art, introducing research and projects on deep learning in pixelization, de-pixelization, and pixel art generation.