arXiv subjects

Pei-Hsuan Tsai

Publications and source records attributed to Pei-Hsuan Tsai.

6 recordsLinked to original sources

arXiv2025

Self-Attention with State-Object Weighted Combination for Compositional Zero Shot Learning

Object recognition has become prevalent across various industries. However, most existing applications are limited to identifying objects alone, without considering their associated states. The ability to recognize both the state and object simultaneously remains less common. One approach to address this is by treating state and object as a single category during training. However, this approach poses challenges in data collection and training since it requires comprehensive data for all possible combinations. Compositional Zero-shot Learning (CZSL) emerges as a viable solution by treating the state and object as distinct categories during training. CZSL facilitates the identification of novel compositions even in the absence of data for every conceivable combination. The current state-of-the-art method, KG-SP, addresses this issue by training distinct classifiers for states and objects, while leveraging a semantic model to evaluate the plausibility of composed compositions. However, KG-SP's accuracy in state and object recognition can be further improved, and it fails to consider the weighting of states and objects during composition. In this study, we propose SASOW, an enhancement of KG-SP that considers the weighting of states and objects while improving composition recognition accuracy. First, we introduce self-attention mechanisms into the classifiers for states and objects, leading to enhanced accuracy in recognizing both. Additionally, we incorporate the weighting of states and objects during composition to generate more reasonable and accurate compositions. Our validation process involves testing SASOW on three established benchmark datasets. Experimental outcomes affirm when compared against OW-CZSL approach, KG-SP, SASOW showcases improvements of 2.1%, 1.7%, and 0.4% in terms of accuracy for unseen compositions across the MIT-States, UT Zappos, and C-GQA datasets, respectively.

cs.CV↗

arXiv2025

Sensitivity-Aware Mixed-Precision Quantization for ReRAM-based Computing-in-Memory

Compute-In-Memory (CIM) systems, particularly those utilizing ReRAM and memristive technologies, offer a promising path toward energy-efficient neural network computation. However, conventional quantization and compression techniques often fail to fully optimize performance and efficiency in these architectures. In this work, we present a structured quantization method that combines sensitivity analysis with mixed-precision strategies to enhance weight storage and computational performance on ReRAM-based CIM systems. Our approach improves ReRAM Crossbar utilization, significantly reducing power consumption, latency, and computational load, while maintaining high accuracy. Experimental results show 86.33% accuracy at 70% compression, alongside a 40% reduction in power consumption, demonstrating the method's effectiveness for power-constrained applications.

cs.AR↗

arXiv2025

Sparse Multi-Modal Transformer with Masking for Alzheimer's Disease Classification

Transformer-based multi-modal intelligent systems often suffer from high computational and energy costs due to dense self-attention, limiting their scalability under resource constraints. This paper presents SMMT, a sparse multi-modal transformer architecture designed to improve efficiency and robustness. Building upon a cascaded multi-modal transformer framework, SMMT introduces cluster-based sparse attention to achieve near linear computational complexity and modality-wise masking to enhance robustness against incomplete inputs. The architecture is evaluated using Alzheimer's Disease classification on the ADNI dataset as a representative multi-modal case study. Experimental results show that SMMT maintains competitive predictive performance while significantly reducing training time, memory usage, and energy consumption compared to dense attention baselines, demonstrating its suitability as a resource-aware architectural component for scalable intelligent systems.

cs.AI↗

arXiv2024

Semantic2Graph: Graph-based Multi-modal Feature Fusion for Action Segmentation in Videos

Video action segmentation have been widely applied in many fields. Most previous studies employed video-based vision models for this purpose. However, they often rely on a large receptive field, LSTM or Transformer methods to capture long-term dependencies within videos, leading to significant computational resource requirements. To address this challenge, graph-based model was proposed. However, previous graph-based models are less accurate. Hence, this study introduces a graph-structured approach named Semantic2Graph, to model long-term dependencies in videos, thereby reducing computational costs and raise the accuracy. We construct a graph structure of video at the frame-level. Temporal edges are utilized to model the temporal relations and action order within videos. Additionally, we have designed positive and negative semantic edges, accompanied by corresponding edge weights, to capture both long-term and short-term semantic relationships in video actions. Node attributes encompass a rich set of multi-modal features extracted from video content, graph structures, and label text, encompassing visual, structural, and semantic cues. To synthesize this multi-modal information effectively, we employ a graph neural network (GNN) model to fuse multi-modal features for node action label classification. Experimental results demonstrate that Semantic2Graph outperforms state-of-the-art methods in terms of performance, particularly on benchmark datasets such as GTEA and 50Salads. Multiple ablation experiments further validate the effectiveness of semantic features in enhancing model performance. Notably, the inclusion of semantic edges in Semantic2Graph allows for the cost-effective capture of long-term dependencies, affirming its utility in addressing the challenges posed by computational resource constraints in video-based vision models.

cs.CV↗

arXiv2022

An Efficient Probe-based Routing for Content-Centric Networking

With the development of new technologies and applications, such as the Internet of Things, smart cities, 5G, and edge computing, traditional Internet Protocol-based (IP-based) networks have been exposed as having many problems. Information-Centric Networking (ICN), Named Data Networking (NDN), and Content-Centric Networking (CCN) are therefore proposed as an alternative for future networks. However, unlike IP-based networks, CCN routing is nondeterministic and difficult to optimize due to frequent in-network caching replacement. This paper presents a novel probe-based routing algorithm that explores real-time in-network caching to ensure the routing table storing the optimal paths to the nearest content provider is up-to-date. Effective probe-selections, Pending Interest Table (PIT) probe, and Forwarding Information Base (FIB) probe are discussed and analyzed by simulation with different performance measurements. Compared with the basic CCN, in terms of qualitative analysis, the additional computational overhead of our approach is O(*) and O(*) on processing interest packets and data packets, respectively. However, in terms of quantitative analysis, our approach reduces the number of timeout interests by 6% and the average response time by 0.6 s. Furthermore, although basic CCN and our approach belong to the same Quality of Service category, our approach outperforms basic CCN in terms of real values. Additionally, our probe-based approach performs better than * and *. Owing to speedup FIB updating by probes, our approach provides more reliable interest packet routing when accounting for router failures. In summary, the results demonstrate that compared to basic CCN, our probe-based routing approach raises FIB accuracy and reduces network congestion and response time, resulting in efficient routing.

cs.NI↗

arXiv2021

A Query-based Routing Table Update Mechanism for Content-Centric Network

Due to the popularity of network applications, such as multimedia, online shopping, Internet of Things (IoT), and 5G, the contents cached in the routers are frequently replaced in Content-Centric Networking (CCN). Generally, cache miss causes numerous propagated packets to get the required content that deteriorates network congestion and delay the response time of consumers. Many caching strategies and routing policies were proposed to solve the problem. This paper presents an alternative solution by designing a query-based routing table update mechanism to increase the accuracy of routing tables. By adding an additional query content in interest packets, our approach real-time explores the cached content in routers and updated the routing table accordingly. This paper uses a general network simulator, ndnSIM, to compare basic CCN and our approach. The results show that our approach improves the response time of consumers and network congestion and is compatible with general forwarding strategies.

cs.NI↗