Alberto Del Bimbo

Alberto Baldrati, Lorenzo Agnolucci, Marco Bertini et al.

Composed Image Retrieval (CIR) aims to retrieve a target image based on a query composed of a reference image and a relative caption that describes the difference between the two images. The high effort and cost required for labeling datasets for CIR hamper the widespread usage of existing methods, as they rely on supervised learning. In this work, we propose a new task, Zero-Shot CIR (ZS-CIR), that aims to address CIR without requiring a labeled training dataset. Our approach, named zero-Shot composEd imAge Retrieval with textuaL invErsion (SEARLE), maps the visual features of the reference image into a pseudo-word token in CLIP token embedding space and integrates it with the relative caption. To support research on ZS-CIR, we introduce an open-domain benchmarking dataset named Composed Image Retrieval on Common Objects in context (CIRCO), which is the first dataset for CIR containing multiple ground truths for each query. The experiments show that SEARLE exhibits better performance than the baselines on the two main datasets for CIR tasks, FashionIQ and CIRR, and on the proposed CIRCO. The dataset, the code and the model are publicly available at https://github.com/miccunifi/SEARLE.

Lorenzo Agnolucci, Leonardo Galteri, Marco Bertini et al.

No-Reference Image Quality Assessment (NR-IQA) aims to develop methods to measure image quality in alignment with human perception without the need for a high-quality reference image. In this work, we propose a self-supervised approach named ARNIQA (leArning distoRtion maNifold for Image Quality Assessment) for modeling the image distortion manifold to obtain quality representations in an intrinsic manner. First, we introduce an image degradation model that randomly composes ordered sequences of consecutively applied distortions. In this way, we can synthetically degrade images with a large variety of degradation patterns. Second, we propose to train our model by maximizing the similarity between the representations of patches of different images distorted equally, despite varying content. Therefore, images degraded in the same manner correspond to neighboring positions within the distortion manifold. Finally, we map the image representations to the quality scores with a simple linear regressor, thus without fine-tuning the encoder weights. The experiments show that our approach achieves state-of-the-art performance on several datasets. In addition, ARNIQA demonstrates improved data efficiency, generalization capabilities, and robustness compared to competing methods. The code and the model are publicly available at https://github.com/miccunifi/ARNIQA.

Alberto Baldrati, Marco Bertini, Tiberio Uricchio et al.

Given a query composed of a reference image and a relative caption, the Composed Image Retrieval goal is to retrieve images visually similar to the reference one that integrates the modifications expressed by the caption. Given that recent research has demonstrated the efficacy of large-scale vision and language pre-trained (VLP) models in various tasks, we rely on features from the OpenAI CLIP model to tackle the considered task. We initially perform a task-oriented fine-tuning of both CLIP encoders using the element-wise sum of visual and textual features. Then, in the second stage, we train a Combiner network that learns to combine the image-text features integrating the bimodal information and providing combined features used to perform the retrieval. We use contrastive learning in both stages of training. Starting from the bare CLIP features as a baseline, experimental results show that the task-oriented fine-tuning and the carefully crafted Combiner network are highly effective and outperform more complex state-of-the-art approaches on FashionIQ and CIRR, two popular and challenging datasets for composed image retrieval. Code and pre-trained models are available at https://github.com/ABaldrati/CLIP4Cir

Giovanni Burbi, Alberto Baldrati, Lorenzo Agnolucci et al.

Multimodal image-text memes are prevalent on the internet, serving as a unique form of communication that combines visual and textual elements to convey humor, ideas, or emotions. However, some memes take a malicious turn, promoting hateful content and perpetuating discrimination. Detecting hateful memes within this multimodal context is a challenging task that requires understanding the intertwined meaning of text and images. In this work, we address this issue by proposing a novel approach named ISSUES for multimodal hateful meme classification. ISSUES leverages a pre-trained CLIP vision-language model and the textual inversion technique to effectively capture the multimodal semantic content of the memes. The experiments show that our method achieves state-of-the-art results on the Hateful Memes Challenge and HarMeme datasets. The code and the pre-trained models are publicly available at https://github.com/miccunifi/ISSUES.

Lorenzo Agnolucci, Leonardo Galteri, Marco Bertini et al.

In the latest years, videoconferencing has taken a fundamental role in interpersonal relations, both for personal and business purposes. Lossy video compression algorithms are the enabling technology for videoconferencing, as they reduce the bandwidth required for real-time video streaming. However, lossy video compression decreases the perceived visual quality. Thus, many techniques for reducing compression artifacts and improving video visual quality have been proposed in recent years. In this work, we propose a novel GAN-based method for compression artifacts reduction in videoconferencing. Given that, in this context, the speaker is typically in front of the camera and remains the same for the entire duration of the transmission, we can maintain a set of reference keyframes of the person from the higher-quality I-frames that are transmitted within the video stream and exploit them to guide the visual quality improvement; a novel aspect of this approach is the update policy that maintains and updates a compact and effective set of reference keyframes. First, we extract multi-scale features from the compressed and reference frames. Then, our architecture combines these features in a progressive manner according to facial landmarks. This allows the restoration of the high-frequency details lost after the video compression. Experiments show that the proposed approach improves visual quality and generates photo-realistic results even with high compression rates. Code and pre-trained networks are publicly available at https://github.com/LorenzoAgnolucci/Keyframes-GAN.

Hondamunige Prasanna Silva, Lorenzo Seidenari, Alberto Del Bimbo

This paper presents a novel reconstruction method that leverages Diffusion Models to protect machine learning classifiers against adversarial attacks, all without requiring any modifications to the classifiers themselves. The susceptibility of machine learning models to minor input perturbations renders them vulnerable to adversarial attacks. While diffusion-based methods are typically disregarded for adversarial defense due to their slow reverse process, this paper demonstrates that our proposed method offers robustness against adversarial threats while preserving clean accuracy, speed, and plug-and-play compatibility. Code at: https://github.com/HondamunigePrasannaSilva/DiffDefence.

Tommaso Barletti, Niccolo' Biondi, Federico Pernici et al.

In this paper, we propose a novel training procedure for the continual representation learning problem in which a neural network model is sequentially learned to alleviate catastrophic forgetting in visual search tasks. Our method, called Contrastive Supervised Distillation (CSD), reduces feature forgetting while learning discriminative features. This is achieved by leveraging labels information in a distillation setting in which the student model is contrastively learned from the teacher model. Extensive experiments show that CSD performs favorably in mitigating catastrophic forgetting by outperforming current state-of-the-art methods. Our results also provide further evidence that feature forgetting evaluated in visual retrieval tasks is not as catastrophic as in classification tasks. Code at: https://github.com/NiccoBiondi/ContrastiveSupervisedDistillation.

Claudio Ferrari, Matteo Serpentoni, Stefano Berretti et al.

Deep learning advanced face recognition to an unprecedented accuracy. However, understanding how local parts of the face affect the overall recognition performance is still mostly unclear. Among others, face swap has been experimented to this end, but just for the entire face. In this paper, we propose to swap facial parts as a way to disentangle the recognition relevance of different face parts, like eyes, nose and mouth. In our method, swapping parts from a source face to a target one is performed by fitting a 3D prior, which establishes dense pixels correspondence between parts, while also handling pose differences. Seamless cloning is then used to obtain smooth transitions between the mapped source regions and the shape and skin tone of the target face. We devised an experimental protocol that allowed us to draw some preliminary conclusions when the swapped images are classified by deep networks, indicating a prominence of the eyes and eyebrows region. Code available at https://github.com/clferrari/FacePartsSwap

Niccolo Biondi, Federico Pernici, Matteo Bruni et al.

In this paper, we propose a method to partially mimic natural intelligence for the problem of lifelong learning representations that are compatible. We take the perspective of a learning agent that is interested in recognizing object instances in an open dynamic universe in a way in which any update to its internal feature representation does not render the features in the gallery unusable for visual search. We refer to this learning problem as Compatible Lifelong Learning Representations (CL2R) as it considers compatible representation learning within the lifelong learning paradigm. We identify stationarity as the property that the feature representation is required to hold to achieve compatibility and propose a novel training procedure that encourages local and global stationarity on the learned representation. Due to stationarity, the statistical properties of the learned features do not change over time, making them interoperable with previously learned features. Extensive experiments on standard benchmark datasets show that our CL2R training procedure outperforms alternative baselines and state-of-the-art methods. We also provide novel metrics to specifically evaluate compatible representation learning under catastrophic forgetting in various sequential learning tasks. Code at https://github.com/NiccoBiondi/CompatibleLifelongRepresentation.

Lorenzo Agnolucci, Leonardo Galteri, Marco Bertini et al.

In this paper, we present a system to restore analog videos of historical archives. These videos often contain severe visual degradation due to the deterioration of their tape supports that require costly and slow manual interventions to recover the original content. The proposed system uses a multi-frame approach and is able to deal with severe tape mistracking, which results in completely scrambled frames. Tests on real-world videos from a major historical video archive show the effectiveness of our demo system. The code and the pre-trained model are publicly available at https://github.com/miccunifi/analog-video-restoration.

Lorenzo Agnolucci, Leonardo Galteri, Marco Bertini et al.

Analog magnetic tapes have been the main video data storage device for several decades. Videos stored on analog videotapes exhibit unique degradation patterns caused by tape aging and reader device malfunctioning that are different from those observed in film and digital video restoration tasks. In this work, we present a reference-based approach for the resToration of digitized Analog videotaPEs (TAPE). We leverage CLIP for zero-shot artifact detection to identify the cleanest frames of each video through textual prompts describing different artifacts. Then, we select the clean frames most similar to the input ones and employ them as references. We design a transformer-based Swin-UNet network that exploits both neighboring and reference frames via our Multi-Reference Spatial Feature Fusion (MRSFF) blocks. MRSFF blocks rely on cross-attention and attention pooling to take advantage of the most useful parts of each reference frame. To address the absence of ground truth in real-world videos, we create a synthetic dataset of videos exhibiting artifacts that closely resemble those commonly found in analog videotapes. Both quantitative and qualitative experiments show the effectiveness of our approach compared to other state-of-the-art methods. The code, the model, and the synthetic dataset are publicly available at https://github.com/miccunifi/TAPE.

Francesco Marchetti, Federico Becattini, Lorenzo Seidenari et al.

Effective modeling of human interactions is of utmost importance when forecasting behaviors such as future trajectories. Each individual, with its motion, influences surrounding agents since everyone obeys to social non-written rules such as collision avoidance or group following. In this paper we model such interactions, which constantly evolve through time, by looking at the problem from an algorithmic point of view, i.e. as a data manipulation task. We present a neural network based on an end-to-end trainable working memory, which acts as an external storage where information about each agent can be continuously written, updated and recalled. We show that our method is capable of learning explainable cause-effect relationships between motions of different agents, obtaining state-of-the-art results on multiple trajectory forecasting datasets.

Roberto Amoroso, Davide Morelli, Marcella Cornia et al.

Recent advancements in diffusion models have enabled the generation of realistic deepfakes from textual prompts in natural language. While these models have numerous benefits across various sectors, they have also raised concerns about the potential misuse of fake images and cast new pressures on fake image detection. In this work, we pioneer a systematic study on deepfake detection generated by state-of-the-art diffusion models. Firstly, we conduct a comprehensive analysis of the performance of contrastive and classification-based visual features, respectively extracted from CLIP-based models and ResNet or ViT-based architectures trained on image classification datasets. Our results demonstrate that fake images share common low-level cues, which render them easily recognizable. Further, we devise a multimodal setting wherein fake images are synthesized by different textual captions, which are used as seeds for a generator. Under this setting, we quantify the performance of fake detection strategies and introduce a contrastive-based disentangling method that lets us analyze the role of the semantics of textual descriptions and low-level perceptual cues. Finally, we release a new dataset, called COCOFake, containing about 1.2M images generated from the original COCO image-caption pairs using two recent text-to-image diffusion models, namely Stable Diffusion v1.4 and v2.0.

Lorenzo Berlincioni, Luca Cultrera, Chiara Albisani et al.

Recently, event cameras have shown large applicability in several computer vision fields especially concerning tasks that require high temporal resolution. In this work, we investigate the usage of such kind of data for emotion recognition by presenting NEFER, a dataset for Neuromorphic Event-based Facial Expression Recognition. NEFER is composed of paired RGB and event videos representing human faces labeled with the respective emotions and also annotated with face bounding boxes and facial landmarks. We detail the data acquisition process as well as providing a baseline method for RGB and event data. The collected data captures subtle micro-expressions, which are hard to spot with RGB data, yet emerge in the event domain. We report a double recognition accuracy for the event-based approach, proving the effectiveness of a neuromorphic approach for analyzing fast and hardly detectable expressions and the emotions they conceal.

Andrea Ciamarra, Roberto Caldelli, Federico Becattini et al.

The ever-increasing use of synthetically generated content in different sectors of our everyday life, one for all media information, poses a strong need for deepfake detection tools in order to avoid the proliferation of altered messages. The process to identify manipulated content, in particular images and videos, is basically performed by looking for the presence of some inconsistencies and/or anomalies specifically due to the fake generation process. Different techniques exist in the scientific literature that exploit diverse ad-hoc features in order to highlight possible modifications. In this paper, we propose to investigate how deepfake creation can impact on the characteristics that the whole scene had at the time of the acquisition. In particular, when an image (video) is captured the overall geometry of the scene (e.g. surfaces) and the acquisition process (e.g. illumination) determine a univocal environment that is directly represented by the image pixel values; all these intrinsic relations are possibly changed by the deepfake generation process. By resorting to the analysis of the characteristics of the surfaces depicted in the image it is possible to obtain a descriptor usable to train a CNN for deepfake detection: we refer to such an approach as SurFake. Experimental results carried out on the FF++ dataset for different kinds of deepfake forgeries and diverse deep learning models confirm that such a feature can be adopted to discriminate between pristine and altered images; furthermore, experiments witness that it can also be combined with visual data to provide a certain improvement in terms of detection accuracy.

Federico Pernici, Matteo Bruni, Claudio Baecchi et al.

Convolutional Neural Networks (CNNs) trained with the Softmax loss are widely used classification models for several vision tasks. Typically, a learnable transformation (i.e. the classifier) is placed at the end of such models returning class scores that are further normalized into probabilities by Softmax. This learnable transformation has a fundamental role in determining the network internal feature representation. In this work we show how to extract from CNNs features with the properties of \emph{maximum} inter-class separability and \emph{maximum} intra-class compactness by setting the parameters of the classifier transformation as not trainable (i.e. fixed). We obtain features similar to what can be obtained with the well-known ``Center Loss'' \cite{wen2016discriminative} and other similar approaches but with several practical advantages including maximal exploitation of the available feature space representation, reduction in the number of network parameters, no need to use other auxiliary losses besides the Softmax. Our approach unifies and generalizes into a common approach two apparently different classes of methods regarding: discriminative features, pioneered by the Center Loss \cite{wen2016discriminative} and fixed classifiers, firstly evaluated in \cite{hoffer2018fix}. Preliminary qualitative experimental results provide some insight on the potentialities of our combined strategy.

Pietro Bongini, Federico Becattini, Alberto Del Bimbo

The use of Deep Learning and Computer Vision in the Cultural Heritage domain is becoming highly relevant in the last few years with lots of applications about audio smart guides, interactive museums and augmented reality. All these technologies require lots of data to work effectively and be useful for the user. In the context of artworks, such data is annotated by experts in an expensive and time consuming process. In particular, for each artwork, an image of the artwork and a description sheet have to be collected in order to perform common tasks like Visual Question Answering. In this paper we propose a method for Visual Question Answering that allows to generate at runtime a description sheet that can be used for answering both visual and contextual questions about the artwork, avoiding completely the image and the annotation process. For this purpose, we investigate on the use of GPT-3 for generating descriptions for artworks analyzing the quality of generated descriptions through captioning metrics. Finally we evaluate the performance for Visual Question Answering and captioning tasks.

Dario Cioni, Lorenzo Berlincioni, Federico Becattini et al.

Cultural heritage applications and advanced machine learning models are creating a fruitful synergy to provide effective and accessible ways of interacting with artworks. Smart audio-guides, personalized art-related content and gamification approaches are just a few examples of how technology can be exploited to provide additional value to artists or exhibitions. Nonetheless, from a machine learning point of view, the amount of available artistic data is often not enough to train effective models. Off-the-shelf computer vision modules can still be exploited to some extent, yet a severe domain shift is present between art images and standard natural image datasets used to train such models. As a result, this can lead to degraded performance. This paper introduces a novel approach to address the challenges of limited annotated data and domain shifts in the cultural heritage domain. By leveraging generative vision-language models, we augment art datasets by generating diverse variations of artworks conditioned on their captions. This augmentation strategy enhances dataset diversity, bridging the gap between natural images and artworks, and improving the alignment of visual cues with knowledge from general-purpose datasets. The generated variations assist in training vision and language models with a deeper understanding of artistic characteristics and that are able to generate better captions with appropriate jargon.

Naima Otberdout, Claudio Ferrari, Mohamed Daoudi et al.

In this work, we address the problem of 4D facial expressions generation. This is usually addressed by animating a neutral 3D face to reach an expression peak, and then get back to the neutral state. In the real world though, people show more complex expressions, and switch from one expression to another. We thus propose a new model that generates transitions between different expressions, and synthesizes long and composed 4D expressions. This involves three sub-problems: (i) modeling the temporal dynamics of expressions, (ii) learning transitions between them, and (iii) deforming a generic mesh. We propose to encode the temporal evolution of expressions using the motion of a set of 3D landmarks, that we learn to generate by training a manifold-valued GAN (Motion3DGAN). To allow the generation of composed expressions, this model accepts two labels encoding the starting and the ending expressions. The final sequence of meshes is generated by a Sparse2Dense mesh Decoder (S2D-Dec) that maps the landmark displacements to a dense, per-vertex displacement of a known mesh topology. By explicitly working with motion trajectories, the model is totally independent from the identity. Extensive experiments on five public datasets show that our proposed approach brings significant improvements with respect to previous solutions, while retaining good generalization to unseen data.

Gabriele Magrini, Federico Becattini, Pietro Pala et al.

In recent years, drone detection has quickly become a subject of extreme interest: the potential for fast-moving objects of contained dimensions to be used for malicious intents or even terrorist attacks has posed attention to the necessity for precise and resilient systems for detecting and identifying such elements. While extensive literature and works exist on object detection based on RGB data, it is also critical to recognize the limits of such modality when applied to UAVs detection. Detecting drones indeed poses several challenges such as fast-moving objects and scenes with a high dynamic range or, even worse, scarce illumination levels. Neuromorphic cameras, on the other hand, can retain precise and rich spatio-temporal information in situations that are challenging for RGB cameras. They are resilient to both high-speed moving objects and scarce illumination settings, while prone to suffer a rapid loss of information when the objects in the scene are static. In this context, we present a novel model for integrating both domains together, leveraging multimodal data to take advantage of the best of both worlds. To this end, we also release NeRDD (Neuromorphic-RGB Drone Detection), a novel spatio-temporally synchronized Event-RGB Drone detection dataset of more than 3.5 hours of multimodal annotated recordings.

Benjamin Szczapa, Mohamed Daoudi, Stefano Berretti et al.

We propose an automatic method to estimate self-reported pain based on facial landmarks extracted from videos. For each video sequence, we decompose the face into four different regions and the pain intensity is measured by modeling the dynamics of facial movement using the landmarks of these regions. A formulation based on Gram matrices is used for representing the trajectory of landmarks on the Riemannian manifold of symmetric positive semi-definite matrices of fixed rank. A curve fitting algorithm is used to smooth the trajectories and temporal alignment is performed to compute the similarity between the trajectories on the manifold. A Support Vector Regression classifier is then trained to encode extracted trajectories into pain intensity levels consistent with self-reported pain intensity measurement. Finally, a late fusion of the estimation for each region is performed to obtain the final predicted pain level. The proposed approach is evaluated on two publicly available datasets, the UNBCMcMaster Shoulder Pain Archive and the Biovid Heat Pain dataset. We compared our method to the state-of-the-art on both datasets using different testing protocols, showing the competitiveness of the proposed approach.

Federico Becattini, Lavinia De Divitiis, Claudio Baecchi et al.

Fashion recommendation is often declined as the task of finding complementary items given a query garment or retrieving outfits that are suitable for a given user. In this work we address the problem by adding an additional semantic layer based on the style of the proposed dressing. We model style according to two important aspects: the mood and the emotion concealed behind color combination patterns and the appropriateness of the retrieved garments for a given type of social event. To address the former we rely on Shigenobu Kobayashi's color image scale, which associated emotional patterns and moods to color triples. The latter instead is analyzed by extracting garments from images of social events. Overall, we integrate in a state of the art garment recommendation framework a style classifier and an event classifier in order to condition recommendation on a given query.

Federico Becattini, Federico Maria Teotini, Alberto Del Bimbo

Suggesting complementary clothing items to compose an outfit is a process of emerging interest, yet it involves a fine understanding of fashion trends and visual aesthetics. Previous works have mainly focused on recommendation by scoring visual appeal and representing garments as ordered sequences or as collections of pairwise-compatible items. This limits the full usage of relations among clothes. We attempt to bridge the gap between outfit recommendation and generation by leveraging a graph-based representation of items in a collection. The work carried out in this paper, tries to build a bridge between outfit recommendation and generation, by discovering new appealing outfits starting from a collection of pre-existing ones. We propose a transformer-based architecture, named TGNN, which exploits multi-headed self attention to capture relations between clothing items in a graph as a message passing step in Convolutional Graph Neural Networks. Specifically, starting from a seed, i.e.~one or more garments, outfit generation is performed by iteratively choosing the garment that is most compatible with the previously chosen ones. Extensive experimentations are conducted with two different datasets, demonstrating the capability of the model to perform seeded outfit generation as well as obtaining state of the art results on compatibility estimation tasks.

Andrea Ciamarra, Federico Becattini, Lorenzo Seidenari et al.

For an autonomous vehicle it is essential to observe the ongoing dynamics of a scene and consequently predict imminent future scenarios to ensure safety to itself and others. This can be done using different sensors and modalities. In this paper we investigate the usage of optical flow for predicting future semantic segmentations. To do so we propose a model that forecasts flow fields autoregressively. Such predictions are then used to guide the inference of a learned warping function that moves instance segmentations on to future frames. Results on the Cityscapes dataset demonstrate the effectiveness of optical-flow methods.

Luca Cultrera, Federico Becattini, Lorenzo Seidenari et al.

Conditional Imitation learning is a common and effective approach to train autonomous driving agents. However, two issues limit the full potential of this approach: (i) the inertia problem, a special case of causal confusion where the agent mistakenly correlates low speed with no acceleration, and (ii) low correlation between offline and online performance due to the accumulation of small errors that brings the agent in a previously unseen state. Both issues are critical for state-aware models, yet informing the driving agent of its internal state as well as the state of the environment is of crucial importance. In this paper we propose a multi-task learning agent based on a multi-stage vision transformer with state token propagation. We feed the state of the vehicle along with the representation of the environment as a special token of the transformer and propagate it throughout the network. This allows us to tackle the aforementioned issues from different angles: guiding the driving policy with learned stop/go information, performing data augmentation directly on the state of the vehicle and visually explaining the model's decisions. We report a drastic decrease in inertia and a high correlation between offline and online metrics.

Lorenzo Agnolucci, Alberto Baldrati, Francesco Todino et al.

Image recognition has recently witnessed a paradigm shift, where vision-language models are now used to perform few-shot classification based on textual prompts. Among these, the CLIP model has shown remarkable capabilities for zero-shot transfer by matching an image and a custom textual prompt in its latent space. This has paved the way for several works that focus on engineering or learning textual contexts for maximizing CLIP's classification capabilities. In this paper, we follow this trend by learning an ensemble of prompts for image classification. We show that learning diverse and possibly shorter contexts improves considerably and consistently the results rather than relying on a single trainable prompt. In particular, we report better few-shot capabilities with no additional cost at inference time. We demonstrate the capabilities of our approach on 11 different benchmarks.

Federico Becattini, Luca Cultrera, Lorenzo Berlincioni et al.

Traditional approaches for analyzing RGB frames are capable of providing a fine-grained understanding of a face from different angles by inferring emotions, poses, shapes, landmarks. However, when it comes to subtle movements standard RGB cameras might fall behind due to their latency, making it hard to detect micro-movements that carry highly informative cues to infer the true emotions of a subject. To address this issue, the usage of event cameras to analyze faces is gaining increasing interest. Nonetheless, all the expertise matured for RGB processing is not directly transferrable to neuromorphic data due to a strong domain shift and intrinsic differences in how data is represented. The lack of labeled data can be considered one of the main causes of this gap, yet gathering data is harder in the event domain since it cannot be crawled from the web and labeling frames should take into account event aggregation rates and the fact that static parts might not be visible in certain frames. In this paper, we first present FACEMORPHIC, a multimodal temporally synchronized face dataset comprising both RGB videos and event streams. The data is labeled at a video level with facial Action Units and also contains streams collected with a variety of applications in mind, ranging from 3D shape estimation to lip-reading. We then show how temporal synchronization can allow effective neuromorphic face analysis without the need to manually annotate videos: we instead leverage cross-modal supervision bridging the domain gap by representing face shapes in a 3D space.

Alberto Baldrati, Marco Bertini, Tiberio Uricchio et al.

Given the recent advances in multimodal image pretraining where visual models trained with semantically dense textual supervision tend to have better generalization capabilities than those trained using categorical attributes or through unsupervised techniques, in this work we investigate how recent CLIP model can be applied in several tasks in artwork domain. We perform exhaustive experiments on the NoisyArt dataset which is a dataset of artwork images crawled from public resources on the web. On such dataset CLIP achieves impressive results on (zero-shot) classification and promising results in both artwork-to-artwork and description-to-artwork domain.

Simone Ricci, Niccolò Biondi, Federico Pernici et al.

Visual retrieval systems face significant challenges when updating models with improved representations due to misalignment between the old and new representations. The costly and resource-intensive backfilling process involves recalculating feature vectors for images in the gallery set whenever a new model is introduced. To address this, prior research has explored backward-compatible training methods that enable direct comparisons between new and old representations without backfilling. Despite these advancements, achieving a balance between backward compatibility and the performance of independently trained models remains an open problem. In this paper, we address it by expanding the representation space with additional dimensions and learning an orthogonal transformation to achieve compatibility with old models and, at the same time, integrate new information. This transformation preserves the original feature space's geometry, ensuring that our model aligns with previous versions while also learning new data. Our Orthogonal Compatible Aligned (OCA) approach eliminates the need for re-indexing during model updates and ensures that features can be compared directly across different model updates without additional mapping functions. Experimental results on CIFAR-100 and ImageNet-1k demonstrate that our method not only maintains compatibility with previous models but also achieves state-of-the-art accuracy, outperforming several existing methods.

Andrea Ciamarra, Federico Becattini, Lorenzo Seidenari et al.

Forecasting motion and spatial positions of objects is of fundamental importance, especially in safety-critical settings such as autonomous driving. In this work, we address the issue by forecasting two different modalities that carry complementary information, namely optical flow and depth. To this end we propose FLODCAST a flow and depth forecasting model that leverages a multitask recurrent architecture, trained to jointly forecast both modalities at once. We stress the importance of training using flows and depth maps together, demonstrating that both tasks improve when the model is informed of the other modality. We train the proposed model to also perform predictions for several timesteps in the future. This provides better supervision and leads to more precise predictions, retaining the capability of the model to yield outputs autoregressively for any future time horizon. We test our model on the challenging Cityscapes dataset, obtaining state of the art results for both flow and depth forecasting. Thanks to the high quality of the generated flows, we also report benefits on the downstream task of segmentation forecasting, injecting our predictions in a flow-based mask-warping framework.

Lorenzo Berlincioni, Stefano Berretti, Marco Bertini et al.

Time varying sequences of 3D point clouds, or 4D point clouds, are now being acquired at an increasing pace in several applications (e.g., LiDAR in autonomous or assisted driving). In many cases, such volume of data is transmitted, thus requiring that proper compression tools are applied to either reduce the resolution or the bandwidth. In this paper, we propose a new solution for upscaling and restoration of time-varying 3D video point clouds after they have been heavily compressed. In consideration of recent growing relevance of 3D applications, %We focused on a model allowing user-side upscaling and artifact removal for 3D video point clouds, a real-time stream of which would require . Our model consists of a specifically designed Graph Convolutional Network (GCN) that combines Dynamic Edge Convolution and Graph Attention Networks for feature aggregation in a Generative Adversarial setting. By taking inspiration PointNet++, We present a different way to sample dense point clouds with the intent to make these modules work in synergy to provide each node enough features about its neighbourhood in order to later on generate new vertices. Compared to other solutions in the literature that address the same task, our proposed model is capable of obtaining comparable results in terms of quality of the reconstruction, while using a substantially lower number of parameters (about 300KB), making our solution deployable in edge computing devices such as LiDAR.

Vittorio Casula, Lorenzo Berlincioni, Luca Cultrera et al.

In the rapidly evolving field of online fashion shopping, the need for more personalized and interactive image retrieval systems has become paramount. Existing methods often struggle with precisely manipulating specific garment attributes without inadvertently affecting others. To address this challenge, we propose GAMMA (Garment Attribute Manipulation with Multi-level Attention), a novel framework that integrates attribute-disentangled representations with a multi-stage attention-based architecture. GAMMA enables targeted manipulation of fashion image attributes, allowing users to refine their searches with high accuracy. By leveraging a dual-encoder Transformer and memory block, our model achieves state-of-the-art performance on popular datasets like Shopping100k and DeepFashion.

Lorenzo Berlincioni, Luca Cultrera, Federico Becattini et al.

This paper investigates the impact of using first names in Large Language Models (LLMs) and Vision Language Models (VLMs), particularly when prompted with ethical decision-making tasks. We propose an approach that appends first names to ethically annotated text scenarios to reveal demographic biases in model outputs. Our study involves a curated list of more than 300 names representing diverse genders and ethnic backgrounds, tested across thousands of moral scenarios. Following the auditing methodologies from social sciences we propose a detailed analysis involving popular LLMs/VLMs to contribute to the field of responsible AI by emphasizing the importance of recognizing and mitigating biases in these systems. Furthermore, we introduce a novel benchmark, the Pratical Scenarios Benchmark (PSB), designed to assess the presence of biases involving gender or demographic prejudices in everyday decision-making scenarios as well as practical scenarios where an LLM might be used to make sensible decisions (e.g., granting mortgages or insurances). This benchmark allows for a comprehensive comparison of model behaviors across different demographic categories, highlighting the risks and biases that may arise in practical applications of LLMs and VLMs.

Alessandro Simoni, Francesco Marchetti, Guido Borghi et al.

Technologies to enable safe and effective collaboration and coexistence between humans and robots have gained significant importance in the last few years. A critical component useful for realizing this collaborative paradigm is the understanding of human and robot 3D poses using non-invasive systems. Therefore, in this paper, we propose a novel vision-based system leveraging depth data to accurately establish the 3D locations of skeleton joints. Specifically, we introduce the concept of Pose Nowcasting, denoting the capability of the proposed system to enhance its current pose estimation accuracy by jointly learning to forecast future poses. The experimental evaluation is conducted on two different datasets, providing accurate and real-time performance and confirming the validity of the proposed method on both the robotic and human scenarios.

Simone Ricci, Niccolò Biondi, Federico Pernici et al.

Minimizing inconsistencies across successive versions of an AI system is as crucial as reducing the overall error. In image classification, such inconsistencies manifest as negative flips, where an updated model misclassifies test samples that were previously classified correctly. This issue becomes increasingly pronounced as the number of training classes grows over time, since adding new categories reduces the margin of each class and may introduce conflicting patterns that undermine their learning process, thereby degrading performance on the original subset. To mitigate negative flips, we propose a novel approach that preserves the margins of the original model while learning an improved one. Our method encourages a larger relative margin between the previously learned and newly introduced classes by introducing an explicit margin-calibration term on the logits. However, overly constraining the logit margin for the new classes can significantly degrade their accuracy compared to a new independently trained model. To address this, we integrate a double-source focal distillation loss with the previous model and a new independently trained model, learning an appropriate decision margin from both old and new data, even under a logit margin calibration. Extensive experiments on image classification benchmarks demonstrate that our approach consistently reduces the negative flip rate with high overall accuracy.

Simone Ricci, Tiberio Uricchio, Alberto Del Bimbo

In this paper, we introduced the novel concept of advisor network to address the problem of noisy labels in image classification. Deep neural networks (DNN) are prone to performance reduction and overfitting problems on training data with noisy annotations. Weighting loss methods aim to mitigate the influence of noisy labels during the training, completely removing their contribution. This discarding process prevents DNNs from learning wrong associations between images and their correct labels but reduces the amount of data used, especially when most of the samples have noisy labels. Differently, our method weighs the feature extracted directly from the classifier without altering the loss value of each data. The advisor helps to focus only on some part of the information present in mislabeled examples, allowing the classifier to leverage that data as well. We trained it with a meta-learning strategy so that it can adapt throughout the training of the main model. We tested our method on CIFAR10 and CIFAR100 with synthetic noise, and on Clothing1M which contains real-world noise, reporting state-of-the-art results.

Alessandra Alfani, Federico Becattini, Lorenzo Seidenari et al.

Several unsupervised and self-supervised approaches have been developed in recent years to learn visual features from large-scale unlabeled datasets. Their main drawback however is that these methods are hardly able to recognize visual features of the same object if it is simply rotated or the perspective of the camera changes. To overcome this limitation and at the same time exploit a useful source of supervision, we take into account video object tracks. Following the intuition that two patches in a track should have similar visual representations in a learned feature space, we adopt an unsupervised clustering-based approach and constrain such representations to be labeled as the same category since they likely belong to the same object or object part. Experimental results on two downstream tasks on different datasets demonstrate the effectiveness of our Online Deep Clustering with Video Track Consistency (ODCT) approach compared to prior work, which did not leverage temporal information. In addition we show that exploiting an unsupervised class-agnostic, yet noisy, track generator yields to better accuracy compared to relying on costly and precise track annotations.

Lorenzo Agnolucci, Alberto Baldrati, Alberto Del Bimbo et al.

Given a query consisting of a reference image and a relative caption, Composed Image Retrieval (CIR) aims to retrieve target images visually similar to the reference one while incorporating the changes specified in the relative caption. The reliance of supervised methods on labor-intensive manually labeled datasets hinders their broad applicability. In this work, we introduce a new task, Zero-Shot CIR (ZS-CIR), that addresses CIR without the need for a labeled training dataset. We propose an approach named iSEARLE (improved zero-Shot composEd imAge Retrieval with textuaL invErsion) that involves mapping the visual information of the reference image into a pseudo-word token in CLIP token embedding space and combining it with the relative caption. To foster research on ZS-CIR, we present an open-domain benchmarking dataset named CIRCO (Composed Image Retrieval on Common Objects in context), the first CIR dataset where each query is labeled with multiple ground truths and a semantic categorization. The experimental results illustrate that iSEARLE obtains state-of-the-art performance on three different CIR datasets -- FashionIQ, CIRR, and the proposed CIRCO -- and two additional evaluation settings, namely domain conversion and object composition. The dataset, the code, and the model are publicly available at https://github.com/miccunifi/SEARLE.

Niccolò Biondi, Federico Pernici, Simone Ricci et al.

Learning compatible representations enables the interchangeable use of semantic features as models are updated over time. This is particularly relevant in search and retrieval systems where it is crucial to avoid reprocessing of the gallery images with the updated model. While recent research has shown promising empirical evidence, there is still a lack of comprehensive theoretical understanding about learning compatible representations. In this paper, we demonstrate that the stationary representations learned by the $d$-Simplex fixed classifier optimally approximate compatibility representation according to the two inequality constraints of its formal definition. This not only establishes a solid foundation for future works in this line of research but also presents implications that can be exploited in practical learning scenarios. An exemplary application is the now-standard practice of downloading and fine-tuning new pre-trained models. Specifically, we show the strengths and critical issues of stationary representations in the case in which a model undergoing sequential fine-tuning is asynchronously replaced by downloading a better-performing model pre-trained elsewhere. Such a representation enables seamless delivery of retrieval service (i.e., no reprocessing of gallery images) and offers improved performance without operational disruptions during model replacement. Code available at: https://github.com/miccunifi/iamcl2r.

Simone Ricci, Niccolò Biondi, Federico Pernici et al.

Retrieval systems rely on representations learned by increasingly powerful models. However, due to the high training cost and inconsistencies in learned representations, there is significant interest in facilitating communication between representations and ensuring compatibility across independently trained neural networks. In the literature, two primary approaches are commonly used to adapt different learned representations: affine transformations, which adapt well to specific distributions but can significantly alter the original representation, and orthogonal transformations, which preserve the original structure with strict geometric constraints but limit adaptability. A key challenge is adapting the latent spaces of updated models to align with those of previous models on downstream distributions while preserving the newly learned representation spaces. In this paper, we impose a relaxed orthogonality constraint, namely $λ$-Orthogonality regularization, while learning an affine transformation, to obtain distribution-specific adaptation while retaining the original learned representations. Extensive experiments across various architectures and datasets validate our approach, demonstrating that it preserves the model's zero-shot performance and ensures compatibility across model updates. Code available at: \href{https://github.com/miccunifi/lambda_orthogonality.git}{https://github.com/miccunifi/lambda\_orthogonality}.

Federico Becattini, Lorenzo Berlincioni, Luca Cultrera et al.

Neuromorphic sensors, also known as event cameras, are a class of imaging devices mimicking the function of biological visual systems. Unlike traditional frame-based cameras, which capture fixed images at discrete intervals, neuromorphic sensors continuously generate events that represent changes in light intensity or motion in the visual field with high temporal resolution and low latency. These properties have proven to be interesting in modeling human faces, both from an effectiveness and a privacy-preserving point of view. Neuromorphic face analysis however is still a raw and unstructured field of research, with several attempts at addressing different tasks with no clear standard or benchmark. This survey paper presents a comprehensive overview of capabilities, challenges and emerging applications in the domain of neuromorphic face analysis, to outline promising directions and open issues. After discussing the fundamental working principles of neuromorphic vision and presenting an in-depth overview of the related research, we explore the current state of available data, standard data representations, emerging challenges, and limitations that require further investigation. This paper aims to highlight the recent process in this evolving field to provide to both experienced and newly come researchers an all-encompassing analysis of the state of the art along with its problems and shortcomings.

Lorenzo Berlincioni, Luca Cultrera, Federico Becattini et al.

Recognizing faces and their underlying emotions is an important aspect of biometrics. In fact, estimating emotional states from faces has been tackled from several angles in the literature. In this paper, we follow the novel route of using neuromorphic data to predict valence and arousal values from faces. Due to the difficulty of gathering event-based annotated videos, we leverage an event camera simulator to create the neuromorphic counterpart of an existing RGB dataset. We demonstrate that not only training models on simulated data can still yield state-of-the-art results in valence-arousal estimation, but also that our trained models can be directly applied to real data without further training to address the downstream task of emotion recognition. In the paper we propose several alternative models to solve the task, both frame-based and video-based.

Luca Cultrera, Federico Becattini, Lorenzo Berlincioni et al.

Face analysis has been studied from different angles to infer emotion, poses, shapes, and landmarks. Traditionally RGB cameras are used, yet for fine-grained tasks standard sensors might not be up to the task due to their latency, making it impossible to record and detect micro-movements that carry a highly informative signal, which is necessary for inferring the true emotions of a subject. Event cameras have been increasingly gaining interest as a possible solution to this and similar high-frame rate tasks. We propose a novel spatiotemporal Vision Transformer model that uses Shifted Patch Tokenization (SPT) and Locality Self-Attention (LSA) to enhance the accuracy of Action Unit classification from event streams. We also address the lack of labeled event data in the literature, which can be considered one of the main causes of an existing gap between the maturity of RGB and neuromorphic vision models. Gathering data is harder in the event domain since it cannot be crawled from the web and labeling frames should take into account event aggregation rates and the fact that static parts might not be visible in certain frames. To this end, we present FACEMORPHIC, a temporally synchronized multimodal face dataset composed of RGB videos and event streams. The dataset is annotated at a video level with facial Action Units and contains streams collected with various possible applications, ranging from 3D shape estimation to lip-reading. We then show how temporal synchronization can allow effective neuromorphic face analysis without the need to manually annotate videos: we instead leverage cross-modal supervision bridging the domain gap by representing face shapes in a 3D space. Our proposed model outperforms baseline methods by effectively capturing spatial and temporal information, crucial for recognizing subtle facial micro-expressions.

Gabriele Magrini, Niccolò Marini, Federico Becattini et al.

Small, fast, and lightweight drones present significant challenges for traditional RGB cameras due to their limitations in capturing fast-moving objects, especially under challenging lighting conditions. Event cameras offer an ideal solution, providing high temporal definition and dynamic range, yet existing benchmarks often lack fine temporal resolution or drone-specific motion patterns, hindering progress in these areas. This paper introduces the Florence RGB-Event Drone dataset (FRED), a novel multimodal dataset specifically designed for drone detection, tracking, and trajectory forecasting, combining RGB video and event streams. FRED features more than 7 hours of densely annotated drone trajectories, using 5 different drone models and including challenging scenarios such as rain and adverse lighting conditions. We provide detailed evaluation protocols and standard metrics for each task, facilitating reproducible benchmarking. The authors hope FRED will advance research in high-speed drone perception and multimodal spatiotemporal understanding.

Gabriele Magrini, Federico Becattini, Luca Cultrera et al.

Event cameras offer significant advantages over traditional frame-based sensors, including higher temporal resolution, lower latency and dynamic range. However, efficiently converting event streams into formats compatible with standard computer vision pipelines remains a challenging problem, particularly in the presence of noise. In this paper, we propose Spike-TBR, a novel event-based encoding strategy based on Temporal Binary Representation (TBR), addressing its vulnerability to noise by integrating spiking neurons. Spike-TBR combines the frame-based advantages of TBR with the noise-filtering capabilities of spiking neural networks, creating a more robust representation of event streams. We evaluate four variants of Spike-TBR, each using different spiking neurons, across multiple datasets, demonstrating superior performance in noise-affected scenarios while improving the results on clean data. Our method bridges the gap between spike-based and frame-based processing, offering a simple noise-resilient solution for event-driven vision applications.

Federico Becattini, Xiaolin Chen, Andrea Puccia et al.

Recommending fashion items often leverages rich user profiles and makes targeted suggestions based on past history and previous purchases. In this paper, we work under the assumption that no prior knowledge is given about a user. We propose to build a user profile on the fly by integrating user reactions as we recommend complementary items to compose an outfit. We present a reinforcement learning agent capable of suggesting appropriate garments and ingesting user feedback so to improve its recommendations and maximize user satisfaction. To train such a model, we resort to a proxy model to be able to simulate having user feedback in the training loop. We experiment on the IQON3000 fashion dataset and we find that a reinforcement learning-based agent becomes capable of improving its recommendations by taking into account personal preferences. Furthermore, such task demonstrated to be hard for non-reinforcement models, that cannot exploit exploration during training.

Yangyang Guo, Liqiang Nie, Harry Cheng et al.

Making each modality in multi-modal data contribute is of vital importance to learning a versatile multi-modal model. Existing methods, however, are often dominated by one or few of modalities during model training, resulting in sub-optimal performance. In this paper, we refer to this problem as modality bias and attempt to study it in the context of multi-modal classification systematically and comprehensively. After stepping into several empirical analysis, we recognize that one modality affects the model prediction more just because this modality has a spurious correlation with instance labels. In order to primarily facilitate the evaluation on the modality bias problem, we construct two datasets respectively for the colored digit recognition and video action recognition tasks in line with the Out-of-Distribution (OoD) protocol. Collaborating with the benchmarks in the visual question answering task, we empirically justify the performance degradation of the existing methods on these OoD datasets, which serves as evidence to justify the modality bias learning. In addition, to overcome this problem, we propose a plug-and-play loss function method, whereby the feature space for each label is adaptively learned according to the training set statistics. Thereafter, we apply this method on eight baselines in total to test its effectiveness. From the results on four datasets regarding the above three tasks, our method yields remarkable performance improvements compared with the baselines, demonstrating its superiority on reducing the modality bias problem.

Niccolo Biondi, Federico Pernici, Matteo Bruni et al.

Compatible features enable the direct comparison of old and new learned features allowing to use them interchangeably over time. In visual search systems, this eliminates the need to extract new features from the gallery-set when the representation model is upgraded with novel data. This has a big value in real applications as re-indexing the gallery-set can be computationally expensive when the gallery-set is large, or even infeasible due to privacy or other concerns of the application. In this paper, we propose CoReS, a new training procedure to learn representations that are \textit{compatible} with those previously learned, grounding on the stationarity of the features as provided by fixed classifiers based on polytopes. With this solution, classes are maximally separated in the representation space and maintain their spatial configuration stationary as new classes are added, so that there is no need to learn any mappings between representations nor to impose pairwise training with the previously learned model. We demonstrate that our training procedure largely outperforms the current state of the art and is particularly effective in the case of multiple upgrades of the training-set, which is the typical case in real applications.

Daniele Mugnai, Federico Pernici, Francesco Turchini et al.

In this paper we exploit Semi-Supervised Learning (SSL) to increase the amount of training data to improve the performance of Fine-Grained Visual Categorization (FGVC). This problem has not been investigated in the past in spite of prohibitive annotation costs that FGVC requires. Our approach leverages unlabeled data with an adversarial optimization strategy in which the internal features representation is obtained with a second-order pooling model. This combination allows to back-propagate the information of the parts, represented by second-order pooling, onto unlabeled data in an adversarial training setting. We demonstrate the effectiveness of the combined use by conducting experiments on six state-of-the-art fine-grained datasets, which include Aircrafts, Stanford Cars, CUB-200-2011, Oxford Flowers, Stanford Dogs, and the recent Semi-Supervised iNaturalist-Aves. Experimental results clearly show that our proposed method has better performance than the only previous approach that examined this problem; it also obtained higher classification accuracy with respect to the supervised learning methods with which we compared.

Francesco Bongini, Lorenzo Berlincioni, Marco Bertini et al.

In this paper we propose a novel data augmentation approach for visual content domains that have scarce training datasets, compositing synthetic 3D objects within real scenes. We show the performance of the proposed system in the context of object detection in thermal videos, a domain where 1) training datasets are very limited compared to visible spectrum datasets and 2) creating full realistic synthetic scenes is extremely cumbersome and expensive due to the difficulty in modeling the thermal properties of the materials of the scene. We compare different augmentation strategies, including state of the art approaches obtained through RL techniques, the injection of simulated data and the employment of a generative model, and study how to best combine our proposed augmentation with these other techniques.Experimental results demonstrate the effectiveness of our approach, and our single-modality detector achieves state-of-the-art results on the FLIR ADAS dataset.