19 citations · 97 across the 49 of their papers we have counts for
50 papers
What Matters for Latent Reasoning with Flow Matching
Yassine Ouali, Adrian Bulat, Georgios Tzimiropoulos
Latent reasoning lets a large language model (LLM) think in a continuous space and verbalize only the answer. We argue that an effective latent thought must meet five requirements:…
UltraViT: Latency-Optimized On-device Vision Encoder for Large Vision-Language Models
Ioannis Maniadis Metaxas, Adrian Bulat, Alberto Baldrati +4
Large Vision-Language Models (LVLMs) remain bottlenecked by massive computational footprints, precluding their deployment on resource-constrained edge devices. While efforts to com…
Hierarchical Image Tokenization for Multi-Scale Image Super Resolution
Isma Hadji, Enrique Sanchez, Adrian Bulat +2
We introduce a multi-scale Image Super Resolution (ISR) method building on recent advances in Visual Auto-Regressive (VAR) modeling. VAR models break image tokenization into additi…
Restore, Assess, Repeat: A Unified Framework for Iterative Image Restoration
I-Hsiang Chen, Isma Hadji, Enrique Sanchez +5
Image restoration aims to recover high quality images from inputs degraded by various factors, such as adverse weather, blur, or low light. While recent studies have shown remarkab…
VISion On Request: Enhanced VLLM efficiency with sparse, dynamically selected, vision-language interactions
Adrian Bulat, Alberto Baldrati, Ioannis Maniadis Metaxas +2
Existing approaches for improving the efficiency of Large Vision-Language Models (LVLMs) are largely based on the concept of visual token reduction. This approach, however, creates…
CycleCap: Improving VLMs Captioning Performance via Self-Supervised Cycle Consistency Fine-Tuning
Marios Krestenitis, Christos Tzelepis, Konstantinos Ioannidis +5
Visual-Language Models (VLMs) have achieved remarkable progress in image captioning, visual question answering, and visual reasoning. Yet they remain prone to vision-language misal…
Deconstructing the Failure of Ideal Noise Correction: A Three-Pillar Diagnosis
Chen Feng, Zhuo Zhi, Zhao Huang +5
Statistically consistent methods based on the noise transition matrix () offer a theoretically grounded solution to Learning with Noisy Labels (LNL), with guarantees of converge…
More Images, More Problems? A Controlled Analysis of VLM Failure Modes
Anurag Das, Adrian Bulat, Alberto Baldrati +4
Large Vision Language Models (LVLMs) have demonstrated remarkable capabilities, yet their proficiency in understanding and reasoning over multiple images remains largely unexplored…
TPSO: Training-Free Diverse Image Generation via Semantic Prompt Embedding Optimization
Debin Meng, Chen Jin, Zheng Gao +3
Image diversity remains a fundamental challenge for text-to-image diffusion models. Low-diversity generation often leads to repetitive outputs, increasing sampling redundancy and h…
Vision-Free Retrieval: Rethinking Multimodal Search with Textual Scene Descriptions
Ioanna Ntinou, Alexandros Xenos, Yassine Ouali +2
Contrastively-trained Vision-Language Models (VLMs), such as CLIP, have become the standard approach for learning discriminative vision-language representations. However, these mod…
Multi-scale Image Super Resolution with a Single Auto-Regressive Model
Enrique Sanchez, Isma Hadji, Adrian Bulat +3
In this paper we tackle Image Super Resolution (ISR), using recent advances in Visual Auto-Regressive (VAR) modeling. VAR iteratively estimates the residual in latent space between…
Fwd2Bot: LVLM Visual Token Compression with Double Forward Bottleneck
Adrian Bulat, Yassine Ouali, Georgios Tzimiropoulos
In this work, we aim to compress the vision tokens of a Large Vision Language Model (LVLM) into a representation that is simultaneously suitable for (a) generative and (b) discrimi…
Edge-SD-SR: Low Latency and Parameter Efficient On-device Super-Resolution with Stable Diffusion via Bidirectional Conditioning
Mehdi Noroozi, Isma Hadji, Victor Escorcia +3
There has been immense progress recently in the visual quality of Stable Diffusion-based Super Resolution (SD-SR). However, deploying large diffusion models on computationally rest…
VladVA: Discriminative Fine-tuning of LVLMs
Yassine Ouali, Adrian Bulat, Alexandros Xenos +4
Contrastively-trained Vision-Language Models (VLMs) like CLIP have become the de facto approach for discriminative vision-language representation learning. However, these models ha…
FAM Diffusion: Frequency and Attention Modulation for High-Resolution Image Generation with Stable Diffusion
Haosen Yang, Adrian Bulat, Isma Hadji +4
Diffusion models are proficient at generating high-quality images. They are however effective only when operating at the resolution used during training. Inference at a scaled reso…
CemiFace: Center-based Semi-hard Synthetic Face Generation for Face Recognition
Zhonglin Sun, Siyang Song, Ioannis Patras +1
Privacy issue is a main concern in developing face recognition techniques. Although synthetic face images can partially mitigate potential legal risks while maintaining effective f…
MM2Latent: Text-to-facial image generation and editing in GANs with multimodal assistance
Debin Meng, Christos Tzelepis, Ioannis Patras +1
Generating human portraits is a hot topic in the image generation area, e.g. mask-to-face generation and text-to-face generation. However, these unimodal generation methods lack co…
CLIPCleaner: Cleaning Noisy Labels with CLIP
Chen Feng, Georgios Tzimiropoulos, Ioannis Patras
Learning with Noisy labels (LNL) poses a significant challenge for the Machine Learning community. Some of the most widely used approaches that select as clean samples for which th…
CLIP-DPO: Vision-Language Models as a Source of Preference for Fixing Hallucinations in LVLMs
Yassine Ouali, Adrian Bulat, Brais Martinez +1
Despite recent successes, LVLMs or Large Vision Language Models are prone to hallucinating details like objects and their properties or relations, limiting their real-world deploym…
MobileQuant: Mobile-friendly Quantization for On-device Language Models
Fuwen Tan, Royson Lee, Łukasz Dudziak +5
Large language models (LLMs) have revolutionized language processing, delivering outstanding results across multiple applications. However, deploying LLMs on edge devices poses sev…
Efficient Unsupervised Visual Representation Learning with Explicit Cluster Balancing
Ioannis Maniadis Metaxas, Georgios Tzimiropoulos, Ioannis Patras
Self-supervised learning has recently emerged as the preeminent pretraining paradigm across and between modalities, with remarkable results. In the image domain specifically, group…
MeMSVD: Long-Range Temporal Structure Capturing Using Incremental SVD
Ioanna Ntinou, Enrique Sanchez, Georgios Tzimiropoulos
This paper is on long-term video understanding where the goal is to recognise human actions over long temporal windows (up to minutes long). In prior work, long temporal context is…
FFF: Fixing Flawed Foundations in contrastive pre-training results in very strong Vision-Language models
Adrian Bulat, Yassine Ouali, Georgios Tzimiropoulos
Despite noise and caption quality having been acknowledged as important factors impacting vision-language contrastive pre-training, in this paper, we show that the full potential o…
VLLMs Provide Better Context for Emotion Understanding Through Common Sense Reasoning
Alexandros Xenos, Niki Maria Foteinopoulou, Ioanna Ntinou +2
Recognising emotions in context involves identifying an individual's apparent emotions while considering contextual cues from the surrounding scene. Previous approaches to this tas…