COINW列出了基於Solana的Meme Coin RFC（延遲查找器硬幣），打開RFC/USDT交易

RFC是由Twitter帳戶@ifindredards發行的Solana區塊鏈建立的模因令牌。以其諷刺評論和參與而聞名

CoinW, a renowned crypto trading platform, has announced the listing of Retard Finder Coin (RFC), a Solana-based meme coin known for its satirical Twitter commentary and community engagement. The exchange will commence trading of the RFC/USDT pair at 1:00 pm (UTC+8) on April 9th. To celebrate the listing of RFC, CoinW is hosting the “RFC Bounty Program” event with a reward pool of $5,000 USDT.

Coinw是一個著名的加密貨幣交易平台，宣布宣布將基於索拉納的Meme Coin列出了雷德發現者硬幣（RFC），以其諷刺的Twitter評論和社區參與而聞名。交易所將於4月9日下午1:00（UTC+8）開始交易RFC/USDT對。為了慶祝RFC的上市，Coinw將舉辦“ RFC賞金計劃”活動，掛池為5,000美元。

Beginning at 5:00 (UTC) on April 9th and concluding at 16:00 (UTC) on April 16th, members of the CoinW community can participate in various events to win a share of the 5,000 USDT prize pool. Registration on the CoinW platform, trading the newly listed RFC/USDT pair, and engaging in community events on Telegram and Twitter will contribute to earning rewards.

從4月9日的5:00（UTC）開始，並於4月16日在16:00（UTC）結束時，Coinw社區的成員可以參加各種活動，以贏得5,000 USDT獎勵池中的一部分。在Coinw平台上進行註冊，交易新列出的RFC/USDT對，並在Telegram和Twitter上參加社區活動會有助於賺取獎勵。

Created by the popular Twitter account @ifindretards, which boasts over 700,000 followers known for its satirical commentary and engagement with a vast community of followers, including frequent interactions with well-known Twitter celebrities. The account has gained immense attention for its humorous takes on cryptocurrency and online culture.

由受歡迎的Twitter帳戶@IfindReDards創建，該帳戶擁有700,000多名追隨者，以其諷刺的評論和與眾多追隨者社區的交往，包括與知名Twitter名人的頻繁互動。該帳戶因其對加密貨幣和在線文化的幽默感而引起了極大的關注。

According to CoinW Research, RFC is a community-driven meme coin without functional utility but has attracted significant attention due to its unique social media narrative. It follows a fair launch model, with 96% of the total supply distributed to the community and only 4% allocated to the developer wallet for liquidity.

根據Coinw Research的說法，RFC是一個沒有功能效用的社區驅動的模因硬幣，但由於其獨特的社交媒體敘事而引起了極大的關注。它遵循公平的發布模型，佔社區總供應量的96％，只有4％分配給開發人員錢包以獲得流動性。

After a successful community vote on April 1st, 2025, to determine the best time for listing, CoinW will be introducing the listing of Retard Finder Coin (RFC).

在2025年4月1日成功進行了社區投票以確定上市的最佳時間之後，Coinw將引入Readard Finder Coin（RFC）的清單。

Applied Sciences Department, Faculty of Science, University of Technology, Malaysia

馬來西亞科技大學科學學院應用科學系

บทความวิชาการ

學術文章

บทความเต็ม (pdf) พร้อม บรรณาธิคมบทความ

文章的完整文章（PDF）

available in

可用

English

英語

บทความนี้

本文

In this paper, we propose a novel approach for blind image watermarking using a hybrid deep learning architecture. The proposed method combines the strengths of convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to embed and extract watermarks in a robust and efficient manner. CNNs are used to extract spatial and spectral features from the cover image, while RNNs are used to model the temporal dependencies among the extracted features. The watermark is then embedded into the cover image using a specially designed embedding module, which minimizes the perceptual distortion of the stamped image. To extract the watermark, a decoder network is designed to recover the watermark bits from the stamped image. Experimental results demonstrate that the proposed method outperforms existing methods in terms of both robustness and imperceptibility. The method is robust to common image processing attacks, such as Gaussian noise, JPEG compression, and scaling. Moreover, the proposed method can achieve high imperceptibility, rendering the embedded watermark invisible to the naked eye.

在本文中，我們提出了一種使用混合深度學習體系結構的新型方法，用於盲圖水印。提出的方法結合了卷積神經網絡（CNN）和復發性神經網絡（RNN）的優勢，以穩健有效的方式嵌入和提取水印。 CNN用於從封面圖像中提取空間和光譜特徵，而RNN用於對提取特徵之間的時間依賴性進行建模。然後，使用特殊設計的嵌入模塊將水印嵌入封面圖像中，從而最大程度地減少了蓋章圖像的感知失真。為了提取水印，設計一個解碼器網絡旨在從蓋章圖像中恢復水印位。實驗結果表明，所提出的方法在魯棒性和不可識別方面都優於現有方法。該方法對常見的圖像處理攻擊具有魯棒性，例如高斯噪聲，JPEG壓縮和縮放。此外，所提出的方法可以實現高度易於識別，從而使嵌入式水印對肉眼看不見。

This paper proposes a novel approach for blind image watermarking using a hybrid deep learning architecture. The proposed method combines the strengths of convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to embed and extract watermarks in a robust and efficient manner. CNNs are used to extract spatial and spectral features from the cover image, while RNNs are used to model the temporal dependencies among the extracted features. The watermark is then embedded into the cover image using a specially designed embedding module, which minimizes the perceptual distortion of the stamped image. To extract the watermark, a decoder network is designed to recover the watermark bits from the stamped image. Experimental results demonstrate that the proposed method outperforms existing methods in terms of both robustness and imperceptibility. The method is robust to common image processing attacks, such as Gaussian noise, JPEG compression, and scaling. Moreover, the proposed method can achieve high imperceptibility, rendering the embedded watermark invisible to the naked eye.

本文提出了一種使用混合深度學習結構的新型方法，用於盲圖水印。提出的方法結合了卷積神經網絡（CNN）和復發性神經網絡（RNN）的優勢，以穩健有效的方式嵌入和提取水印。 CNN用於從封面圖像中提取空間和光譜特徵，而RNN用於對提取特徵之間的時間依賴性進行建模。然後，使用特殊設計的嵌入模塊將水印嵌入封面圖像中，從而最大程度地減少了蓋章圖像的感知失真。為了提取水印，設計一個解碼器網絡旨在從蓋章圖像中恢復水印位。實驗結果表明，所提出的方法在魯棒性和不可識別方面都優於現有方法。該方法對常見的圖像處理攻擊具有魯棒性，例如高斯噪聲，JPEG壓縮和縮放。此外，所提出的方法可以實現高度易於識別，從而使嵌入式水印對肉眼看不見。

Image watermarking is an important technique for protecting digital content. It involves embedding a watermark signal into a cover image to identify the copyright holder or track the usage of the image. The watermark should be robust to common image processing attacks, such as Gaussian noise, JPEG compression, and scaling. At the same time, the watermark should be imperceptible to avoid affecting the visual quality of the image.

圖像水印是保護數字內容的重要技術。它涉及將水印信號嵌入封面圖像中，以識別版權持有人或跟踪圖像的使用情況。水印應適用於常見的圖像處理攻擊，例如高斯噪聲，JPEG壓縮和縮放。同時，應避免影響圖像的視覺質量的水印。

Deep learning has achieved promising results in various low-level vision tasks, such as image denoising, super-resolution, and image manipulation detection. Recently, deep learning methods have also been applied to image watermarking. Convolutional neural networks (CNNs) are good at extracting spatial and spectral features from images, while recurrent neural networks (RNNs) are suitable for modeling temporal dependencies among data.

深度學習在各種低級視覺任務中取得了令人鼓舞的結果，例如圖像降解，超分辨率和圖像操縱檢測。最近，深度學習方法也已應用於圖像水印。卷積神經網絡（CNN）擅長從圖像中提取空間和光譜特徵，而復發性神經網絡（RNN）適合在數據之間建模時間依賴性。

In this paper, we propose a hybrid deep learning architecture for blind image watermarking, which combines the strengths of CNNs and RNNs. CNNs are used to extract features from the cover image, and RNNs are used to embed the watermark into the extracted features. A specially designed embedding module is proposed to minimize the perceptual distortion of the stamped image. To extract the

在本文中，我們提出了用於盲圖水印的混合深度學習結構，結合了CNN和RNN的優勢。 CNN用於從封面圖像中提取特徵，RNN用於將水印嵌入到提取的特徵中。提出了一個專門設計的嵌入模塊，以最大程度地減少郵票圖像的感知失真。提取