Temporal action localization aims to identify the boundaries of the action of interest in a video. Most existing methods take a two-stage approach: first, identify a set of action proposals; then, based on this set, determine the accurate temporal locations of the action of interest. However, the diversely distributed semantics of a video over time have not been well considered, which could compromise the localization performance, especially for ubiquitous short actions or events (e.g., a fall in healthcare and a traffic violation in surveillance). To address this problem, we propose a novel deep learning architecture, namely an adaptive template-guided self-attention network, to characterize the proposals adaptively with their relevant frames. An input video is segmented into temporal frames, within which the spatio-temporal patterns are formulated by a global–Local Transformer-based encoder. Each frame is associated with a number of proposals of different lengths as their starting frame. Learnable templates for proposals of different lengths are introduced, and each template guides the sampling for proposals with a specific length. It formulates the probabilities for a proposal to form the representation of certain spatio-temporal patterns from its relevant temporal frames. Therefore, the semantics of a proposal can be formulated in an adaptive manner, and a feature map of all proposals can be appropriately characterized. To estimate the IoU of these proposals with ground truth actions, a two-level scheme is introduced. A shortcut connection is also utilized to refine the predictions by using the convolutions of the feature map from coarse to fine. Comprehensive experiments on two benchmark datasets demonstrate the state-of-the-art performance of our proposed method: 32.6% mAP@IoU 0.7 on THUMOS-14 and 9.35% mAP@IoU 0.95 on ActivityNet-1.3.
Internet of Things (IoT) devices are frequent targets of cyber attacks, and firmware emulation is essential for enabling automated techniques like fuzzing to uncover vulnerabilities. However, the growing use of encryption in firmware poses significant challenges for building accurate emulated environments. To overcome this, we present EncryptAFL, the first IoT fuzzer specifically tailored for devices with encrypted firmware. In contrast to existing methods that primarily focus on peripheral emulation, EncryptAFL places greater emphasis on automating the decryption process. We evaluate EncryptAFL using real-world firmware and known vulnerabilities, demonstrating its effectiveness in both emulation and vulnerability detection.
Transportation records cross organizations with different urgency, privacy, and audit requirements. Encrypted databases support local storage and queries, while ledgers can preserve commitments that other organizations verify. This paper presents trafficBCDB, which binds an ordered record batch to its policy and ledger-selection context and returns authenticated range proofs with ledger receipts. The implementation uses threshold proxy re-encryption, three SQLite replicas, and separate four-node Geth Clique and Tendermint networks. The routing study compares nine policies using controlled workloads and actual New York City taxi pickup records, with shared observations and 20 paired random seeds. On the generated workload, trafficBCDB reaches 98.184% joint constraint satisfaction at 303.4 ms modeled latency, compared with 96.803% for a calibrated fixed policy. A constrained optimizer reaches 98.191% at 300.9 ms. On the citywide taxi replay, trafficBCDB reaches 26.64% under the stated synthetic ledger and policy assumptions. Separate native experiments measure inclusion, ledger-specific acceptance, encrypted storage, proof verification, and decryption. The results establish a concrete accountable data path and expose its routing, confidentiality, and availability limits.