Skip to content
View Luigina2001's full-sized avatar
9 followers · 9 following

Achievements

Achievement: Pair ExtraordinaireAchievement: Pull Shark

Achievements

Achievement: Pair ExtraordinaireAchievement: Pull Shark

Block or report Luigina2001

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Luigina2001/README.md

Hi there, I'm Luigina!

I am a Computer Science graduate specialized in Data Science & AI at the University of Salerno. 🎓
I combine a researcher's mindset with practical engineering to build impactful AI systems, with a strong focus on Computer Vision and Deepfake Detection.

Gmail LinkedIn Latest paper

🔬 Featured research

Depthtective

A depth-aware framework for spatio-temporal deepfake detection
Published in Array (Elsevier), vol. 32, 2026

Depthtective detects manipulated videos using only two adjacent frames: the absolute residuals of the aligned RGB frames and of their monocular depth maps are fused into a compact four-channel tensor, classified by a standard CNN. A contrastive learning variant reaches an AUC of 97.26% on Celeb-DF and 98.35% on DFDC.

Paper Code

 Depthtective pipeline

📄 Publications & presentations

  • C. Bisogni, L. Costante, V. Loia, M. Nappi, C. Pero. Depthtective: A depth-aware framework for spatio-temporal deepfake detection. Array (Elsevier), 32, 101278, 2026. [DOI] [Code]
  • C. Bisogni, L. Costante, A. Miglino, C. Pero. Lightweight Head Pose Estimation for Deepfake Detection: A Performance Benchmark on FF++. Presented at the SAILS 2025 workshop, held in conjunction with the 28th European Conference on Artificial Intelligence (ECAI 2025).

Research interests: spatio-temporal deepfake detection · biometrics · depth estimation · representation learning

💻 Selected projects

🧬 C3DP

Contrastive 3D Protein Prediction
Multi-modal framework linking protein sequences to their 3D structures, combining DNABERT-2 and GNNs trained with the InfoNCE loss.

PyTorch Geometric Transformers Contrastive Learning

🧠 GiBERTino

Dialogue Graph Modeling
Models multi-party dialogues as graphs for discourse parsing: EDUs are encoded with ModernBERT and context is propagated with GNNs.

PyTorch Geometric HuggingFace NLP

👾 Ipomoea Aquatica

Reinforcement Learning on Space Invaders
Benchmark of DQN, A3C and Q-Learning agents on the Atari game, comparing performance and training efficiency.

OpenAI Gym TensorFlow Reinforcement Learning

💬 NaLA

Sentiment Analysis on Bluesky
Sentiment analysis of Bluesky posts comparing BERTweet, RoBERTa, BiRNN and Naive Bayes, with a focus on the accuracy/inference-cost trade-off.

Transformers NLP Model Efficiency

🛠️ Tech Stack & Skills

Core AI & Data Languages Dev Tools & DB

Domains: Computer Vision • NLP • Bioinformatics • Deepfake Detection • Social Network Analysis

Pinned Loading

  1. Depthtective Depthtective Public

    Official PyTorch implementation of "Depthtective: A Depth-Aware Framework for Spatio-Temporal Deepfake Detection" (Array, 2026). Detects manipulated videos from the RGB and depth residuals of just …

    Python

  2. contrastive-3d-protein-prediction contrastive-3d-protein-prediction Public

    Forked from angelonazzaro/contrastive-3d-protein-prediction

    A contrastive approach for 3d protein prediction

    Python 1

  3. GiBERTino GiBERTino Public

    Supervised extraction of sub-dialogues from multi-part dialogues

    Python 1

  4. SMiLe-CoDe SMiLe-CoDe Public

    Heuristic seed-set selection for Majority Cascade in social networks

    Python