This SHREC 2025 track dedicated to protein surface shape retrieval involved 9 participating teams. We evaluated the performance in retrieval of 15 proposed methods on a large dataset of 11,555 protein surfaces with calculated electrostatic potential (a key molecular surface descriptor). The performance in retrieval of the proposed methods was evaluated through different metrics (Accuracy, Balanced accuracy, F1 score, Precision and Recall). The best retrieval performance was achieved by the proposed methods that used the electrostatic potential complementary to molecular surface shape. This observation was also valid for classes with limited data which highlights the importance of taking into account additional molecular surface descriptors.

Urban environments present significant challenges to commercial Global Navigation Satellite Systems (GNSS) receivers due to degraded satellite visibility and Non-line-of-sight (NLOS) receptions. Mitigating NLOS receptions for GNSS is essential, especially for safety-critical and reliability-critical location-based applications. Traditional physical error channel propagation modeling encountered bottlenecks since the NLOS and multipath errors cannot be modeled accurately in complex urban environments. Data-driven methods show significant potential for effectively classifying GNSS Line-of-sight (LOS) and NLOS. This paper proposes the CarNet - a generative Convolutional Neural Network (CNN)-based GNSS LOS/NLOS classifier by transforming multivariate time-series data into images. CarNet comprises two modules: an image generator and an image classifier. The image generator enriches and augments the original 1-dimension feature vector into 2-dimension feature maps and the image classifier uses an inception-based CNN to realize multi-scale feature extraction and classification. The proposed architecture is trained and tested on more than 6 h of real vehicle data collected in different challenging environments (about 1.6 million samples). A thorough benchmark is conducted, comparing CarNet against the existing mainstream Artificial Intelligence (AI) methods. The results with cross-validation on unseen data indicate that CarNet achieves the highest accuracy, i.e., 81.47% while maintaining the optimal balance between precision for both classes: 83.3% for LOS and 70.99% for NLOS. Finally, positioning accuracy is assessed using a reweighting strategy based on the LOS/NLOS information predicted by CarNet. The assessment of total datasets shows that CarNet weighting can achieve the best accuracy compared to the traditional weighting schemes based on signal-to-noise ratio or satellite elevation. CarNet shows strong potential for embedding into GNSS receivers to enhance positioning accuracy in complex urban environments, benefiting a wide range of location-based applications such as autonomous driving , emergency response, and urban logistics.

Speech recognition system converts audio into texts by utilizing deep learning algorithms. Numerous works have demonstrated various adversarial example (AE) attacks, i.e., adding carefully-crafted noises can trick the speech recognition system into outputting completely incorrect texts. This paper aims to reveal the distinctive properties of adversarial audio in terms of phonetics. We believe analyzing the distinctive properties is critical in understanding adversarial attacks on ASR models, as well as guiding the generation and defense of AEs. Thus, we aim to answer three questions: (1) What are the distinctive properties of adversarial audio that are common to diverse attacks? (2) How to quantify these distinctive properties? (3) How can we use these properties to improve the security of ASR models? To answer these questions, we perform a large-scale measurement based on acoustic features and statistical analysis. By measuring a total of 612,000 acoustic-statistical feature vectors for 2,400 audio samples, we obtain four insights on the distinctive properties, i.e., filling energy gap, speech-like morphology, disordered signal, and abnormal linguistic pattern. Based on these properties, we design a naturalness score to assess the stealthiness of attacks and propose an adversarial example detector with an average accuracy of 91.1%.

Gesture recognition is a tool to enable novel interactions with different techniques and applications, like Mixed Reality and Virtual Reality environments. With all the recent advancements in gesture recognition from skeletal data, it is still unclear how well state-of-the-art techniques perform in a scenario using precise motions with two hands. This paper presents the results of the SHREC 2024 contest organized to evaluate methods for their recognition of highly similar hand motions using the skeletal spatial coordinate data of both hands. The task is the recognition of 7 motion classes given their spatial coordinates in a frame-by-frame motion. The skeletal data has been captured using a Vicon system and pre-processed into a coordinate system using Blender and Vicon Shogun Post. We created a small, novel dataset with a high variety of durations in frames. This paper shows the results of the contest, showing the techniques created by the 5 research groups on this challenging task and comparing them to our baseline method.

Due to the open nature of voice and voice interface, an adversary can spoof voice recognition systems by replaying pre-recorded voice commands from legitimate users, known as the voice replay attack. Existing detection methods against voice replay attacks mainly rely on extra hardware to determine the sound source or require excessive computing resources to train a classifier with abundant acoustic features. In this paper, we propose Anti-Replay, a fast and lightweight detection system for voice replay attacks. To overcome the challenge of redundant classification features and complex calculation, we first investigate the time-frequency spectrum difference between the genuine human voice and the replayed audio caused by the non-linear distortion of the attacker’s microphones and speakers. Then, we design 5 types with a total of 77 features in both the time and frequency domains and propose a convolutional neural network classifier SE-ResNet50 for attack detection. Evaluations against the datasets of ASVspoof2017, ASVspoof2019, and ASVspoof2021 demonstrate that Anti-Replay can achieve an average equal error rate (EER) of 1.36% across three datasets. Meanwhile, Anti-Replay decreases the training time by 52.3% and 90.2% and decreases the model size by 83.5% and 99.9% compared with the baseline model CQCC-GMM and the state-of-the-art method Res2Net. We have also confirmed that our system is effective in detecting the adaptive replay attack.

A voiceprint is the distinctive pattern of human voices widely used for authentication in voice assistants. This paper investigates the impact of speech contents and speech rates on the distinctiveness of voiceprint, and has obtained answers to three questions by studying 2457 speakers and 21,500,000 test samples: 1) What are the influential factors that users can control to affect the distinctiveness of voiceprints? 2) How to quantify the distinctiveness for given speeches, e.g., the speech of wake-up words when activating voice assistants? 3) How to help users select wake-up words and adjust the speech rate to improve distinctiveness levels? To answer those questions, we break down speeches into phones, and experimentally obtain the correlation between false recognition rates and the richness, order, length, and elements of the phones. Then, we define the PROLE Score that can reflect the voice distinctiveness, and evaluate 30 wake-up words of 19 commercial voice assistant products to provide recommendations on selecting secure voiceprint words. We also measure the correlation between false recognition rates and speech rates, and define the TER Score that reveals the distance of distinctiveness from the secure voiceprint, and it guides users to adjust their speech rate to a secure value.

In this article, we propose a new approach for the cow heat detection from endoscopic images. Our approach permits to identify on the fly the cow heat state through two successive stages, namely cervix detection then heat classification. For this purpose, images are analyzed by a Transformer based detection model to localize the cervix, in which case they are analyzed by a CNN-based heat classification model. The proposed approach permits to assist the farmer during the insemination operation by localizing the cervix in an accurate way. Moreover, the confidence level of the final decision of the classification model is increased by focusing its analysis only on cervix images. The effectiveness of our method is demonstrated on our generated dataset and the obtained performance outperform the state of the art.