Gioacchino Tangari

Gioacchino Tangari, Shreesh Keskar, Hassan Jameel Asghar et al.

Biometric authentication service providers often claim that it is not possible to reverse-engineer a user's raw biometric sample, such as a fingerprint or a face image, from its mathematical (feature-space) representation. In this paper, we investigate this claim on the specific example of deep neural network (DNN) embeddings. Inversion of DNN embeddings has been investigated for explaining deep image representations or synthesizing normalized images. Existing studies leverage full access to all layers of the original model, as well as all possible information on the original dataset. For the biometric authentication use case, we need to investigate this under adversarial settings where an attacker has access to a feature-space representation but no direct access to the exact original dataset nor the original learned model. Instead, we assume varying degree of attacker's background knowledge about the distribution of the dataset as well as the original learned model (architecture and training process). In these cases, we show that the attacker can exploit off-the-shelf DNN models and public datasets, to mimic the behaviour of the original learned model to varying degrees of success, based only on the obtained representation and attacker's prior knowledge. We propose a two-pronged attack that first infers the original DNN by exploiting the model footprint on the embedding, and then reconstructs the raw data by using the inferred model. We show the practicality of the attack on popular DNNs trained for two prominent biometric modalities, face and fingerprint recognition. The attack can effectively infer the original recognition model (mean accuracy 83\% for faces, 86\% for fingerprints), and can craft effective biometric reconstructions that are successfully authenticated with 1-vs-1 authentication accuracy of up to 92\% for some models.

Cong Duy Vu Hoang, Gioacchino Tangari, Clemence Lanfranchi et al.

The growing adoption of large language models (LLMs) in business applications has amplified interest in Natural Language to SQL (NL2SQL) solutions, in which there is competing demand for high performance and efficiency. Domain- and customer-specific requirements further complicate the problem. To address this conundrum, we introduce Distill-C, a distilled customization framework tailored for NL2SQL tasks. Distill-C utilizes large teacher LLMs to produce high-quality synthetic data through a robust and scalable pipeline. Finetuning smaller and open-source LLMs on this synthesized data enables them to rival or outperform teacher models an order of magnitude larger. Evaluated on multiple challenging benchmarks, Distill-C achieves an average improvement of 36% in execution accuracy compared to the base models from three distinct LLM families. Additionally, on three internal customer benchmarks, Distill-C demonstrates a 22.6% performance improvement over the base models. Our results demonstrate that Distill-C is an effective, high-performing and generalizable approach for deploying lightweight yet powerful NL2SQL models, delivering exceptional accuracies while maintaining low computational cost.

Meng Chen, Philip Arthur, Qianyu Feng et al.

Large language models (LLMs) have shown impressive performance in \emph{code} understanding and generation, making coding tasks a key focus for researchers due to their practical applications and value as a testbed for LLM evaluation. Data synthesis and filtering techniques have been widely adopted and shown to be highly effective in this context. In this paper, we present a focused survey and taxonomy of these techniques, emphasizing recent advancements. We highlight key challenges, explore future research directions, and offer practical guidance for new researchers entering the field.

Saad Sajid Hashmi, Nazar Waheed, Gioacchino Tangari et al.

Contemporary mobile applications (apps) are designed to track, use, and share users' data, often without their consent, which results in potential privacy and transparency issues. To investigate whether mobile apps have always been (non-)transparent regarding how they collect information about users, we perform a longitudinal analysis of the historical versions of 268 Android apps. These apps comprise 5,240 app releases or versions between 2008 and 2016. We detect inconsistencies between apps' behaviors and the stated use of data collection in privacy policies to reveal compliance issues. We utilize machine learning techniques for the classification of the privacy policy text to identify the purported practices that collect and/or share users' personal information, such as phone numbers and email addresses. We then uncover the data leaks of an app through static and dynamic analysis. Over time, our results show a steady increase in the number of apps' data collection practices that are undisclosed in the privacy policies. This behavior is particularly troubling since privacy policy is the primary tool for describing the app's privacy protection practices. We find that newer versions of the apps are likely to be more non-compliant than their preceding versions. The discrepancies between the purported and the actual data practices show that privacy policies are often incoherent with the apps' behaviors, thus defying the 'notice and choice' principle when users install apps.