In text-to-image diffusion models, the cross-attention map of each text token indicates the specific image regions attended. Comparing these maps of syntactically related tokens provides insights into how well the generated image reflects the text prompt. For example, in the prompt, "a black car and a white clock", the cross-attention maps for "black" and "car" should focus on overlapping regions to depict a black car, while "car" and "clock" should not. Incorrect overlapping in the maps generally produces generation flaws such as missing objects and incorrect attribute binding. Our study makes the key observations investigating this issue in the existing text-to-image models: (1) the similarity in text embeddings between different tokens—used as conditioning inputs—can cause their cross-attention maps to focus on the same image regions; and (2) text embeddings often fail to faithfully capture syntactic relations already within text attention maps. As a result, such syntactic relationships can be overlooked in cross-attention module, leading to inaccurate image generation. To address this, we propose a method that directly transfers syntactic relations from the text attention maps to the cross-attention module via a test-time optimization. Our approach leverages this inherent yet unexploited information within text attention maps to enhance image-text semantic alignment across diverse prompts, without relying on external guidance. Our project page and code are available at: https://t-sam-diffusion.github.io/

Botnet detection is an active research topic as botnets are a source of many malicious activities, including distributed denial-of-service (DDoS), click-fraud, spamming, and crypto-mining attacks. However, it is getting more complicated to identify botnets due to the continuous evolution of botnet software and families that harness new types of devices and attack vectors. Recent studies employing machine learning (ML) showed improved performance to detect botnets to some extent, but they are still limited and ineffective with the lack of sequential pattern analysis, which is a key to detect various classes of botnets. In this paper, we propose a novel botnet detection method, built upon Recurrent Variational Autoencoder (RVAE), that effectively captures sequential characteristics of botnet anomalies. We validate the feasibility of the proposed method with the CTU-13 dataset that have been widely employed for botnet detection studies, and show that our method is at least comparable to existing techniques in terms of detection accuracy. In addition, our experimental results show that the proposed method can detect previously unseen botnets by utilizing sequential patterns of network traffic. We will also show how our method can detect botnets in the streaming mode, which is the essential requirement to perform real-time, on-line detection.

In automated material handling systems (AMHS), delivery time is an important issue directly associated with the production cost and the quality of the product. In this paper, we propose a dynamic routing strategy to shorten delivery time and delay. We set the target of control by analyzing traffic flows and selecting the region with the highest flow rate and congestion frequency. Then, we impose a routing cost in order to dynamically reflect the real-time changes of traffic states. Our deep reinforcement learning model consists of a Q-learning step and a recurrent neural network, through which traffic states and action values are predicted. Experiment results show that the proposed method decreases manufacturing costs while increasing productivity. Additionally, we find evidence the reinforcement learning structure proposed in this study can autonomously and dynamically adjust to the changes in traffic patterns.

The factors impacting on what users purchase can be classified as an internal factor which is time-invariant user's unique taste and an external factor which is time-varying item characteristic. However, the current recommendation system has a limitation of making recommendations based only on the user's history without taking into account the item features trends of the time, which prevents precise recommendations. In this paper, the recommendation system that reflects inter-items trends of time-based bin is proposed. We focus on creating a hybrid recommender system that could effectively combine time-varying content data with the users history data. Specifically, we use Conditional Variational Autoencoder (VAE) to add a time dynamic item features to user-item implicit feedback data. In this case, distributed representation of items in the specific period is used as a condition that is added to input and latent variable of VAE respectively. In detail, the distributed representation per time bin can be extracted using LSTM. By applying a condition into VAE, a hybrid recommendation system can be created to reflect the item time-varying features. The proposed model in this paper differs from current studies in that it reflects the changing characteristics inherent of the products and utilizes it for recommendation. The Movielens 1M data and Amazon women's clothing dataset are used for the evaluation of the proposed model.