Marius Leordeanu

Andrada Gobeaja, Ionut Hodoroaga, Elena Burceanu et al.

Can a dataset be recognized from the spurious correlations it induces during training? We argue that datasets leave dataset-specific traces in a model's learned semantic correlation structure: incidental regularities that are predictive within a dataset, but not causal for the underlying task, can be internalized during training. We use this insight to study dataset-level membership inference, moving beyond existing methods that rely on behavioral or distributional evidence such as confidence scores, losses, margins, generated samples, or query responses. We introduce a white-box semantic fingerprinting approach based on semantic correlation descriptors (SCDs), which capture the semantic correlation structure learned by a model and make it comparable across dataset mixtures. In a controlled leave-one-dataset-out diagnostic, SCDs recover dataset-specific changes and perfectly separate matching from non-matching dataset pairs. We then propose a practical SCD-based membership score that tests whether a target dataset is part of a model's training mixture using only the model's SCD and the target dataset's standalone SCD, without requiring leave-one-dataset-out models. Across three diverse experimental settings, with dataset groups for natural language inference, emotion classification, and medical text classification, we test both the advantages and limitations of SCD-based membership inference with different degrees of semantic separation and keyword support between dataset splits. On average, the classifier based on this score achieves the highest performance and the lowest std, outperforming black-box baselines RMIA, Attack-P, and LiRA, as well as the white-box SIF baseline. These results show that dataset membership can be traced through internal semantic correlations, with the largest relative gain exceeding 60% in ROC-AUC when dataset groups expose distinct semantic particularities.

Alexandra Budisteanu, Dragos Costea, Alina Marcu et al.

The task of generating novel views of real scenes is increasingly important nowadays when AI models become able to create realistic new worlds. In many practical applications, it is important for novel view synthesis methods to stay grounded in the physical world as much as possible, while also being able to imagine it from previously unseen views. While most current methods are developed and tested in virtual environments with small scenes and no errors in pose and depth information, we push the boundaries to the real-world domain of large scales in the new context of UAVs. Our algorithmic contributions are two folds. First, we manage to stay anchored in the real 3D world, by introducing an efficient multi-scale voxel carving method, which is able to accommodate significant noises in pose, depth, and illumination variations, while being able to reconstruct the view of the world from drastically different poses at test time. Second, our final high-resolution output is efficiently self-trained on data automatically generated by the voxel carving module, which gives it the flexibility to adapt efficiently to any scene. We demonstrated the effectiveness of our method on highly complex and large-scale scenes in real environments while outperforming the current state-of-the-art. Our code is publicly available: https://github.com/onorabil/MSVC.

Alina Marcu, Mihai Pirvu, Dragos Costea et al.

We present a method for learning multiple scene representations given a small labeled set, by exploiting the relationships between such representations in the form of a multi-task hypergraph. We also show how we can use the hypergraph to improve a powerful pretrained VisTransformer model without any additional labeled data. In our hypergraph, each node is an interpretation layer (e.g., depth or segmentation) of the scene. Within each hyperedge, one or several input nodes predict the layer at the output node. Thus, each node could be an input node in some hyperedges and an output node in others. In this way, multiple paths can reach the same node, to form ensembles from which we obtain robust pseudolabels, which allow self-supervised learning in the hypergraph. We test different ensemble models and different types of hyperedges and show superior performance to other multi-task graph models in the field. We also introduce Dronescapes, a large video dataset captured with UAVs in different complex real-world scenes, with multiple representations, suitable for multi-task learning.

Mihai Masala, Marius Leordeanu, Mihai Dascalu et al.

Vision-Language Models (VLMs) largely follow the text-only LLM trajectory, excelling on English benchmarks but sharply degrading on low-resource languages, where neither large-scale image-text corpora nor culturally grounded evaluations exist. We present a systematic study of building a language-specific VLM for Romanian, covering the full pipeline from data construction to architectural choices. We translate established English VLM training and evaluation corpora into Romanian, applying machine translation to textual annotations and to in-image text, preserving visual grounding while adapting the textual content. Using this data, we train and ablate a series of VLMs to isolate the contribution of (i) vision backbones of varying scale and pretraining, (ii) language backbones from multilingual to Romanian-adapted LLMs, and (iii) OCR-style image-text data. We further curate HoraVQA, a culturally native evaluation set grounded in Romanian everyday scenes. Romanian-adapted VLMs consistently outperform their same-sized counterparts and, across all evaluated benchmarks, even surpass models from the next larger size category.

Florin Condrea, Saikiran Rapaka, Lucian Itu et al.

Pulmonary Embolisms (PE) represent a leading cause of cardiovascular death. While medical imaging, through computed tomographic pulmonary angiography (CTPA), represents the gold standard for PE diagnosis, it is still susceptible to misdiagnosis or significant diagnosis delays, which may be fatal for critical cases. Despite the recently demonstrated power of deep learning to bring a significant boost in performance in a wide range of medical imaging tasks, there are still very few published researches on automatic pulmonary embolism detection. Herein we introduce a deep learning based approach, which efficiently combines computer vision and deep neural networks for pulmonary embolism detection in CTPA. Our method features novel improvements along three orthogonal axes: 1) automatic detection of anatomical structures; 2) anatomical aware pretraining, and 3) a dual-hop deep neural net for PE detection. We obtain state-of-the-art results on the publicly available multicenter large-scale RSNA dataset.

Mihai Pirvu, Alina Marcu, Alexandra Dobrescu et al.

There are many ways of interpreting the world and they are highly interdependent. We exploit such complex dependencies and introduce a powerful multi-task hypergraph, in which every node is a task and different paths through the hypergraph reaching a given task become unsupervised teachers, by forming ensembles that learn to generate reliable pseudolabels for that task. Each hyperedge is part of an ensemble teacher for a given task and it is also a student of the self-supervised hypergraph system. We apply our model to one of the most important problems of our times, that of Earth Observation, which is highly multi-task and it often suffers from missing ground-truth data. By performing extensive experiments on the NASA NEO Dataset, spanning a period of 22 years, we demonstrate the value of our multi-task semi-supervised approach, by consistent improvements over strong baselines and recent work. We also show that the hypergraph can adapt unsupervised to gradual data distribution shifts and reliably recover, through its multi-task self-supervision process, the missing data for several observational layers for up to seven years.

Lucian Bicsi, Bogdan Alexe, Radu Tudor Ionescu et al.

We propose JEDI, a multi-dataset semi-supervised learning method, which efficiently combines knowledge from multiple experts, learned on different datasets, to train and improve the performance of individual, per dataset, student models. Our approach achieves this by addressing two important problems in current machine learning research: generalization across datasets and limitations of supervised training due to scarcity of labeled data. We start with an arbitrary number of experts, pretrained on their own specific dataset, which form the initial set of student models. The teachers are immediately derived by concatenating the feature representations from the penultimate layers of the students. We then train all models in a student-teacher semi-supervised learning scenario until convergence. In our efficient approach, student-teacher training is carried out jointly and end-to-end, showing that both students and teachers improve their generalization capacity during training. We validate our approach on four video action recognition datasets. By simultaneously considering all datasets within a unified semi-supervised setting, we demonstrate significant improvements over the initial experts.

Mihai Masala, Nicolae Cudlenco, Traian Rebedea et al.

Artificial Intelligence makes great advances today and starts to bridge the gap between vision and language. However, we are still far from understanding, explaining and controlling explicitly the visual content from a linguistic perspective, because we still lack a common explainable representation between the two domains. In this work we come to address this limitation and propose the Graph of Events in Space and Time (GEST), by which we can represent, create and explain, both visual and linguistic stories. We provide a theoretical justification of our model and an experimental validation, which proves that GEST can bring a solid complementary value along powerful deep learning models. In particular, GEST can help improve at the content-level the generation of videos from text, by being easily incorporated into our novel video generation engine. Additionally, by using efficient graph matching techniques, the GEST graphs can also improve the comparisons between texts at the semantic level.

Elena Burceanu, Marius Leordeanu

We pose video object segmentation as spectral graph clustering in space and time, with one graph node for each pixel and edges forming local space-time neighborhoods. We claim that the strongest cluster in this video graph represents the salient object. We start by introducing a novel and efficient method based on 3D filtering for approximating the spectral solution, as the principal eigenvector of the graph's adjacency matrix, without explicitly building the matrix. This key property allows us to have a fast parallel implementation on GPU, orders of magnitude faster than classical approaches for computing the eigenvector. Our motivation for a spectral space-time clustering approach, unique in video semantic segmentation literature, is that such clustering is dedicated to preserving object consistency over time, which we evaluate using our novel segmentation consistency measure. Further on, we show how to efficiently learn the solution over multiple input feature channels. Finally, we extend the formulation of our approach beyond the segmentation task, into the realm of object tracking. In extensive experiments we show significant improvements over top methods, as well as over powerful ensembles that combine them, achieving state-of-the-art on multiple benchmarks, both for tracking and segmentation.

Nicolae Cudlenco, Mihai Masala, Marius Leordeanu

Existing multi-agent video generation systems use LLM agents to orchestrate neural video generators, producing visually impressive but semantically unreliable outputs with no ground truth annotations. We present an agentic system that inverts this paradigm: instead of generating pixels, the LLM constructs a formal Graph of Events in Space and Time (GEST) -- a structured specification of actors, actions, objects, and temporal constraints -- which is then executed deterministically in a 3D game engine. A staged LLM refinement pipeline fails entirely at this task (0 of 50 attempts produce an executable specification), motivating a fundamentally different architecture based on a separation of concerns: the LLM handles narrative planning through natural language reasoning, while a programmatic state backend enforces all simulator constraints through validated tool calls, guaranteeing that every generated specification is executable by construction. The system uses a hierarchical two-agent architecture -- a Director that plans the story and a Scene Builder that constructs individual scenes through a round-based state machine -- with dedicated Relation Subagents that populate the logical and semantic edge types of the GEST formalism that procedural generation leaves empty, making this the first approach to exercise the full expressive capacity of the representation. We evaluate in two stages: autonomous generation against procedural baselines via a 3-model LLM jury, where agentic narratives win 79% of text and 74% of video comparisons; and seeded generation where the same text is given to our system, VEO 3.1, and WAN 2.2, with human annotations showing engine-generated videos substantially outperform neural generators on physical validity (58% vs 25% and 20%) and semantic alignment (3.75/5 vs 2.33 and 1.50).

Dragos Costea, Alina Marcu, Cristina Lazar et al.

We study the ongoing debate regarding the statistical fidelity of AI-generated data compared to human-generated data in the context of non-verbal communication using full body motion. Concretely, we ask if contemporary generative models move beyond surface mimicry to participate in the silent, but expressive dialogue of body language. We tackle this question by introducing the first framework that generates a natural non-verbal interaction between Human and AI in real-time from 2D body keypoints. Our experiments utilize four lightweight architectures which run at up to 100 FPS on an NVIDIA Orin Nano, effectively closing the perception-action loop needed for natural Human-AI interaction. We trained on 437 human video clips and demonstrated that pretraining on synthetically-generated sequences reduces motion errors significantly, without sacrificing speed. Yet, a measurable reality gap persists. When the best model is evaluated on keypoints extracted from cutting-edge text-to-video systems, such as SORA and VEO, we observe that performance drops on SORA-generated clips. However, it degrades far less on VEO, suggesting that temporal coherence, not image fidelity, drives real-world performance. Our results demonstrate that statistically distinguishable differences persist between Human and AI motion.

Nicolae Cudlenco, Mihai Masala, Marius Leordeanu

Generating complex multi-actor scenario videos remains difficult even for state-of-the-art neural generators, while evaluating them is hard due to the lack of ground truth for physical plausibility and semantic faithfulness. We introduce GTASA, a corpus of multi-actor videos with per-frame spatial relation graphs and event-level temporal mappings, and the system that produced it based on Graphs of Events in Space and Time (GEST): GEST-Engine. We compare our method with both open and closed source neural generators and prove both qualitatively (human evaluation of physical validity and semantic alignment) and quantitatively (via training video captioning models) the clear advantages of our method. Probing four frozen video encoders across 11 spatiotemporal reasoning tasks enabled by GTASA's exact 3D ground truth reveals that self-supervised encoders encode spatial structure significantly better than VLM visual encoders.

Mihai-Cristian Pîrvu, Marius Leordeanu

The real-world is inherently multi-modal at its core. Our tools observe and take snapshots of it, in digital form, such as videos or sounds, however much of it is lost. Similarly for actions and information passing between humans, languages are used as a written form of communication. Traditionally, Machine Learning models have been unimodal (i.e. rgb -> semantic or text -> sentiment_class). Recent trends go towards bi-modality, where images and text are learned together, however, in order to truly understand the world, we need to integrate all these independent modalities. In this work we try to combine as many visual modalities as we can using little to no human supervision. In order to do this, we use pre-trained experts and procedural combinations between them on top of raw videos using a fully autonomous data-pipeline, which we also open-source. We then make use of PHG-MAE, a model specifically designed to leverage multi-modal data. We show that this model which was efficiently distilled into a low-parameter (<1M) can have competitive results compared to models of ~300M parameters. We deploy this model and analyze the use-case of real-time semantic segmentation from handheld devices or webcams on commodity hardware. Finally, we deploy other off-the-shelf models using the same framework, such as DPT for near real-time depth estimation.

Daniel Airinei, Elena Burceanu, Marius Leordeanu

Indoor navigation is a difficult task, as it generally comes with poor GPS access, forcing solutions to rely on other sources of information. While significant progress continues to be made in this area, deployment to production applications is still lacking, given the complexity and additional requirements of current solutions. Here, we introduce an efficient, real-time and easily deployable deep learning approach, based on visual input only, that can predict the direction towards a target from images captured by a mobile device. Our technical approach, based on a novel graph-based path generation method, combined with explainable data augmentation and curriculum learning, includes contributions that make the process of data collection, annotation and training, as automatic as possible, efficient and robust. On the practical side, we introduce a novel largescale dataset, with video footage inside a relatively large shopping mall, in which each frame is annotated with the correct next direction towards different specific target destinations. Different from current methods, ours relies solely on vision, avoiding the need of special sensors, additional markers placed along the path, knowledge of the scene map or internet access. We also created an easy to use application for Android, which we plan to make publicly available. We make all our data and code available along with visual demos on our project site

Mihai Masala, Denis C. Ilie-Ablachim, Alexandru Dima et al.

In recent years, Large Language Models (LLMs) have achieved almost human-like performance on various tasks. While some LLMs have been trained on multilingual data, most of the training data is in English; hence, their performance in English greatly exceeds other languages. To our knowledge, we are the first to collect and translate a large collection of texts, instructions, and benchmarks and train, evaluate, and release open-source LLMs tailored for Romanian. We evaluate our methods on four different categories, including academic benchmarks, MT-Bench (manually translated), and a professionally built historical, cultural, and social benchmark adapted to Romanian. We argue for the usefulness and high performance of RoLLMs by obtaining state-of-the-art results across the board. We publicly release all resources (i.e., data, training and evaluation code, models) to support and encourage research on Romanian LLMs while concurrently creating a generalizable recipe, adequate for other low or less-resourced languages.

Emanuela Haller, Elena Burceanu, Marius Leordeanu

The human ability to synchronize the feedback from all their senses inspired recent works in multi-task and multi-modal learning. While these works rely on expensive supervision, our multi-task graph requires only pseudo-labels from expert models. Every graph node represents a task, and each edge learns between tasks transformations. Once initialized, the graph learns self-supervised, based on a novel consensus shift algorithm that intelligently exploits the agreement between graph pathways to generate new pseudo-labels for the next learning cycle. We demonstrate significant improvement from one unsupervised learning iteration to the next, outperforming related recent methods in extensive multi-task learning experiments on two challenging datasets. Our code is available at https://github.com/bit-ml/cshift.

Sebastian-Ion Nae, Mihai-Eugen Barbu, Sebastian Mocanu et al.

Autonomous agents such as indoor drones must learn new object classes in real-time while limiting catastrophic forgetting, motivating Class-Incremental Learning (CIL). However, most unmanned aerial vehicle (UAV) datasets focus on outdoor scenes and offer limited temporally coherent indoor videos. We introduce an indoor dataset of $14,400$ frames capturing inter-drone and ground vehicle footage, annotated via a semi-automatic workflow with a $98.6\%$ first-pass labeling agreement before final manual verification. Using this dataset, we benchmark 3 replay-based CIL strategies: Experience Replay (ER), Maximally Interfered Retrieval (MIR), and Forgetting-Aware Replay (FAR), using YOLOv11-nano as a resource-efficient detector for deployment-constrained UAV platforms. Under tight memory budgets ($5-10\%$ replay), FAR performs better than the rest, achieving an average accuracy (ACC, $mAP_{50-95}$ across increments) of $82.96\%$ with $5\%$ replay. Gradient-weighted class activation mapping (Grad-CAM) analysis shows attention shifts across classes in mixed scenes, which is associated with reduced localization quality for drones. The experiments further demonstrate that replay-based continual learning can be effectively applied to edge aerial systems. Overall, this work contributes an indoor UAV video dataset with preserved temporal coherence and an evaluation of replay-based CIL under limited replay budgets. Project page: https://spacetime-vision-robotics-laboratory.github.io/learning-on-the-fly-cl

Ciprian Constantinescu, Marius Leordeanu

Humans effortlessly identify objects by leveraging a rich understanding of the surrounding scene, including spatial relationships, material properties, and the co-occurrence of other objects. In contrast, most computational object recognition systems operate on isolated image regions, devoid of meaning in isolation, thus ignoring this vital contextual information. This paper argues for the critical role of context and introduces a novel framework for contextual object classification. We first construct a Geo-Semantic Contextual Graph (GSCG) from a single monocular image. This rich, structured representation is built by integrating a metric depth estimator with a unified panoptic and material segmentation model. The GSCG encodes objects as nodes with detailed geometric, chromatic, and material attributes, and their spatial relationships as edges. This explicit graph structure makes the model's reasoning process inherently interpretable. We then propose a specialized graph-based classifier that aggregates features from a target object, its immediate neighbors, and the global scene context to predict its class. Through extensive ablation studies, we demonstrate that our context-aware model achieves a classification accuracy of 73.4%, dramatically outperforming context-agnostic versions (as low as 38.4%). Furthermore, our GSCG-based approach significantly surpasses strong baselines, including fine-tuned ResNet models (max 53.5%) and a state-of-the-art multimodal Large Language Model (LLM), Llama 4 Scout, which, even when given the full image alongside a detailed description of objects, maxes out at 42.3%. These results on COCO 2017 train/val splits highlight the superiority of explicitly structured and interpretable context for object recognition tasks.

Mihai Masala, Denis C. Ilie-Ablachim, Dragos Corlatescu et al.

In recent years, Large Language Models (LLMs) have achieved almost human-like performance on various tasks. While some LLMs have been trained on multilingual data, most of the training data is in English. Hence, their performance in English greatly exceeds their performance in other languages. This document presents our approach to training and evaluating the first foundational and chat LLM specialized for Romanian.

Florin Condrea, Saikiran Rapaka, Marius Leordeanu

Pulmonary Embolism (PE) are a leading cause of cardiovascular death. Computed tomographic pulmonary angiography (CTPA) is the gold standard for PE diagnosis, with growing interest in AI-based diagnostic assistance. However, these algorithms are limited by scarce fine-grained annotations of thromboembolic burden. We address this challenge with iExplain, a weakly supervised learning algorithm that transforms coarse image-level annotations into detailed pixel-level PE masks through iterative model explainability. Our approach generates soft segmentation maps used to mask detected regions, enabling the process to repeat and discover additional embolisms that would be missed in a single pass. This iterative refinement effectively captures complete PE regions and detects multiple distinct embolisms. Models trained on these automatically generated annotations achieve excellent PE detection performance, with significant improvements at each iteration. We demonstrate iExplain's effectiveness on the RSPECT augmented dataset, achieving results comparable to strongly supervised methods while outperforming existing weakly supervised methods.

Mihai Masala, Marius Leordeanu

In the current era of Machine Learning, Transformers have become the de facto approach across a variety of domains, such as computer vision and natural language processing. Transformer-based solutions are the backbone of current state-of-the-art methods for language generation, image and video classification, segmentation, action and object recognition, among many others. Interestingly enough, while these state-of-the-art methods produce impressive results in their respective domains, the problem of understanding the relationship between vision and language is still beyond our reach. In this work, we propose a common ground between vision and language based on events in space and time in an explainable and programmatic way, to connect learning-based vision and language state of the art models and provide a solution to the long standing problem of describing videos in natural language. We validate that our algorithmic approach is able to generate coherent, rich and relevant textual descriptions on videos collected from a variety of datasets, using both standard metrics (e.g. Bleu, ROUGE) and the modern LLM-as-a-Jury approach.

Dragos Costea, Alina Marcu, Cristina Lazar et al.

Modeling face-to-face communication in computer vision, which focuses on recognizing and analyzing nonverbal cues and behaviors during interactions, serves as the foundation for our proposed alternative to text-based Human-AI interaction. By leveraging nonverbal visual communication, through facial expressions, head and body movements, we aim to enhance engagement and capture the user's attention through a novel improvisational element, that goes beyond mirroring gestures. Our goal is to track and analyze facial expressions, and other nonverbal cues in real-time, and use this information to build models that can predict and understand human behavior. Operating in real-time and requiring minimal computational resources, our approach signifies a major leap forward in making AI interactions more natural and accessible. We offer three different complementary approaches, based on retrieval, statistical, and deep learning techniques. A key novelty of our work is the integration of an artistic component atop an efficient human-computer interaction system, using art as a medium to transmit emotions. Our approach is not art-specific and can be adapted to various paintings, animations, and avatars. In our experiments, we compare state-of-the-art diffusion models as mediums for emotion translation in 2D, and our 3D avatar, Maia, that we introduce in this work, with not just facial movements but also body motions for a more natural and engaging experience. We demonstrate the effectiveness of our approach in translating AI-generated emotions into human-relatable expressions, through both human and automatic evaluation procedures, highlighting its potential to significantly enhance the naturalness and engagement of Human-AI interactions across various applications.

Sebastian Mocanu, Emil Slusanschi, Marius Leordeanu

This paper presents a vision-only autonomous flight system for small UAVs operating in controlled indoor environments. The system combines semantic segmentation with monocular depth estimation to enable obstacle avoidance, scene exploration, and autonomous safe landing operations without requiring GPS or expensive sensors such as LiDAR. A key innovation is an adaptive scale factor algorithm that converts non-metric monocular depth predictions into accurate metric distance measurements by leveraging semantic ground plane detection and camera intrinsic parameters, achieving a mean distance error of 14.4 cm. The approach uses a knowledge distillation framework where a color-based Support Vector Machine (SVM) teacher generates training data for a lightweight U-Net student network (1.6M parameters) capable of real-time semantic segmentation. For more complex environments, the SVM teacher can be replaced with a state-of-the-art segmentation model. Testing was conducted in a controlled 5x4 meter laboratory environment with eight cardboard obstacles simulating urban structures. Extensive validation across 30 flight tests in a real-world environment and 100 flight tests in a digital-twin environment demonstrates that the combined segmentation and depth approach increases the distance traveled during surveillance and reduces mission time while maintaining 100% success rates. The system is further optimized through end-to-end learning, where a compact student neural network learns complete flight policies from demonstration data generated by our best-performing method, achieving an 87.5% autonomous mission success rate. This work advances practical vision-based drone navigation in structured environments, demonstrating solutions for metric depth estimation and computational efficiency challenges that enable deployment on resource-constrained platforms.

Sebastian Mocanu, Sebastian-Ion Nae, Mihai-Eugen Barbu et al.

This work introduces a self-supervised neuro-analytical, cost efficient, model for visual-based quadrotor control in which a small 1.7M parameters student ConvNet learns automatically from an analytical teacher, an improved image-based visual servoing (IBVS) controller. Our IBVS system solves numerical instabilities by reducing the classical visual servoing equations and enabling efficient stable image feature detection. Through knowledge distillation, the student model achieves 11x faster inference compared to the teacher IBVS pipeline, while demonstrating similar control accuracy at a significantly lower computational and memory cost. Our vision-only self-supervised neuro-analytic control, enables quadrotor orientation and movement without requiring explicit geometric models or fiducial markers. The proposed methodology leverages simulation-to-reality transfer learning and is validated on a small drone platform in GPS-denied indoor environments. Our key contributions include: (1) an analytical IBVS teacher that solves numerical instabilities inherent in classical approaches, (2) a two-stage segmentation pipeline combining YOLOv11 with a U-Net-based mask splitter for robust anterior-posterior vehicle segmentation to correctly estimate the orientation of the target, and (3) an efficient knowledge distillation dual-path system, which transfers geometric visual servoing capabilities from the analytical IBVS teacher to a compact and small student neural network that outperforms the teacher, while being suitable for real-time onboard deployment.

Mihai Masala, Marius Leordeanu

The task of describing video content in natural language is commonly referred to as video captioning. Unlike conventional video captions, which are typically brief and widely available, long-form paragraph descriptions in natural language are scarce. This limitation of current datasets is due to the expensive human manual annotation required and to the highly challenging task of explaining the language formation process from the perspective of the underlying story, as a complex system of interconnected events in space and time. Through a thorough analysis of recently published methods and available datasets, we identify a general lack of published resources dedicated to the problem of describing videos in complex language, beyond the level of descriptions in the form of enumerations of simple captions. Furthermore, while state-of-the-art methods produce impressive results on the task of generating shorter captions from videos by direct end-to-end learning between the videos and text, the problem of explaining the relationship between vision and language is still beyond our reach. In this work, we propose a shared representation between vision and language, based on graphs of events in space and time, which can be obtained in an explainable and analytical way, to integrate and connect multiple vision tasks to produce the final natural language description. Moreover, we also demonstrate how our automated and explainable video description generation process can function as a fully automatic teacher to effectively train direct, end-to-end neural student pathways, within a self-supervised neuro-analytical system. We validate that our explainable neuro-analytical approach generates coherent, rich and relevant textual descriptions on videos collected from multiple varied datasets, using both standard evaluation metrics, human annotations and consensus from ensembles of state-of-the-art VLMs.

Andrei Jelea, Ahmed Nabil Belbachir, Marius Leordeanu

We introduce Generalized Test-Time Augmentation (GTTA), a highly effective method for improving the performance of a trained model, which unlike other existing Test-Time Augmentation approaches from the literature is general enough to be used off-the-shelf for many vision and non-vision tasks, such as classification, regression, image segmentation and object detection. By applying a new general data transformation, that randomly perturbs multiple times the PCA subspace projection of a test input, GTTA forms robust ensembles at test time in which, due to sound statistical properties, the structural and systematic noises in the initial input data is filtered out and final estimator errors are reduced. Different from other existing methods, we also propose a final self-supervised learning stage in which the ensemble output, acting as an unsupervised teacher, is used to train the initial single student model, thus reducing significantly the test time computational cost, at no loss in accuracy. Our tests and comparisons to strong TTA approaches and SoTA models on various vision and non-vision well-known datasets and tasks, such as image classification and segmentation, speech recognition and house price prediction, validate the generality of the proposed GTTA. Furthermore, we also prove its effectiveness on the more specific real-world task of salmon segmentation and detection in low-visibility underwater videos, for which we introduce DeepSalmon, the largest dataset of its kind in the literature.

Andrei Jelea, Ahmed Nabil Belbachir, Marius Leordeanu

Solving fish segmentation in underwater videos, a real-world problem of great practical value in marine and aquaculture industry, is a challenging task due to the difficulty of the filming environment, poor visibility and limited existing annotated underwater fish data. In order to overcome these obstacles, we introduce a novel two stage unsupervised segmentation approach that requires no human annotations and combines artificially created and real images. Our method generates challenging synthetic training data, by placing virtual fish in real-world underwater habitats, after performing fish transformations such as Thin Plate Spline shape warping and color Histogram Matching, which realistically integrate synthetic fish into the backgrounds, making the generated images increasingly closer to the real world data with every stage of our approach. While we validate our unsupervised method on the popular DeepFish dataset, obtaining a performance close to a fully-supervised SoTA model, we further show its effectiveness on the specific case of salmon segmentation in underwater videos, for which we introduce DeepSalmon, the largest dataset of its kind in the literature (30 GB). Moreover, on both datasets we prove the capability of our approach to boost the performance of the fully-supervised SoTA model.

Dragos Costea, Alina Marcu, Marius Leordeanu

Generating novel views from recorded videos is crucial for enabling autonomous UAV navigation. Recent advancements in neural rendering have facilitated the rapid development of methods capable of rendering new trajectories. However, these methods often fail to generalize well to regions far from the training data without an optimized flight path, leading to suboptimal reconstructions. We propose a self-supervised cyclic neural-analytic pipeline that combines high-quality neural rendering outputs with precise geometric insights from analytical methods. Our solution improves RGB and mesh reconstructions for novel view synthesis, especially in undersampled areas and regions that are completely different from the training dataset. We use an effective transformer-based architecture for image reconstruction to refine and adapt the synthesis process, enabling effective handling of novel, unseen poses without relying on extensive labeled datasets. Our findings demonstrate substantial improvements in rendering views of novel and also 3D reconstruction, which to the best of our knowledge is a first, setting a new standard for autonomous navigation in complex outdoor environments.

Alexandru-Raul Todoran, Marius Leordeanu

There is an increasing number of real-world problems in computer vision and machine learning requiring to take into consideration multiple interpretation layers (modalities or views) of the world and learn how they relate to each other. For example, in the case of Earth Observations from satellite data, it is important to be able to predict one observation layer (e.g. vegetation index) from other layers (e.g. water vapor, snow cover, temperature etc), in order to best understand how the Earth System functions and also be able to reliably predict information for one layer when the data is missing (e.g. due to measurement failure or error).

Mihai Masala, Nicolae Cudlenco, Traian Rebedea et al.

One of the essential human skills is the ability to seamlessly build an inner representation of the world. By exploiting this representation, humans are capable of easily finding consensus between visual, auditory and linguistic perspectives. In this work, we set out to understand and emulate this ability through an explicit representation for both vision and language - Graphs of Events in Space and Time (GEST). GEST alows us to measure the similarity between texts and videos in a semantic and fully explainable way, through graph matching. It also allows us to generate text and videos from a common representation that provides a well understood content. In this work we show that the graph matching similarity metrics based on GEST outperform classical text generation metrics and can also boost the performance of state of art, heavily trained metrics.

Ioana Croitoru, Simion-Vlad Bogolin, Marius Leordeanu et al.

In recent years, considerable progress on the task of text-video retrieval has been achieved by leveraging large-scale pretraining on visual and audio datasets to construct powerful video encoders. By contrast, despite the natural symmetry, the design of effective algorithms for exploiting large-scale language pretraining remains under-explored. In this work, we are the first to investigate the design of such algorithms and propose a novel generalized distillation method, TeachText, which leverages complementary cues from multiple text encoders to provide an enhanced supervisory signal to the retrieval model. Moreover, we extend our method to video side modalities and show that we can effectively reduce the number of used modalities at test time without compromising performance. Our approach advances the state of the art on several video retrieval benchmarks by a significant margin and adds no computational overhead at test time. Last but not least, we show an effective application of our method for eliminating noise from retrieval datasets. Code and data can be found at https://www.robots.ox.ac.uk/~vgg/research/teachtext/.

Emanuela Haller, Adina Magda Florea, Marius Leordeanu

We propose a dual system for unsupervised object segmentation in video, which brings together two modules with complementary properties: a space-time graph that discovers objects in videos and a deep network that learns powerful object features. The system uses an iterative knowledge exchange policy. A novel spectral space-time clustering process on the graph produces unsupervised segmentation masks passed to the network as pseudo-labels. The net learns to segment in single frames what the graph discovers in video and passes back to the graph strong image-level features that improve its node-level features in the next iteration. Knowledge is exchanged for several cycles until convergence. The graph has one node per each video pixel, but the object discovery is fast. It uses a novel power iteration algorithm computing the main space-time cluster as the principal eigenvector of a special Feature-Motion matrix without actually computing the matrix. The thorough experimental analysis validates our theoretical claims and proves the effectiveness of the cyclical knowledge exchange. We also perform experiments on the supervised scenario, incorporating features pretrained with human supervision. We achieve state-of-the-art level on unsupervised and supervised scenarios on four challenging datasets: DAVIS, SegTrack, YouTube-Objects, and DAVSOD.

Marius Leordeanu, Mihai Pirvu, Dragos Costea et al.

We address the challenging problem of semi-supervised learning in the context of multiple visual interpretations of the world by finding consensus in a graph of neural networks. Each graph node is a scene interpretation layer, while each edge is a deep net that transforms one layer at one node into another from a different node. During the supervised phase edge networks are trained independently. During the next unsupervised stage edge nets are trained on the pseudo-ground truth provided by consensus among multiple paths that reach the nets' start and end nodes. These paths act as ensemble teachers for any given edge and strong consensus is used for high-confidence supervisory signal. The unsupervised learning process is repeated over several generations, in which each edge becomes a "student" and also part of different ensemble "teachers" for training other students. By optimizing such consensus between different paths, the graph reaches consistency and robustness over multiple interpretations and generations, in the face of unknown labels. We give theoretical justifications of the proposed idea and validate it on a large dataset. We show how prediction of different representations such as depth, semantic segmentation, surface normals and pose from RGB input could be effectively learned through self-supervised consensus in our graph. We also compare to state-of-the-art methods for multi-task and semi-supervised learning and show superior performance.

Alina Marcu, Vlad Licaret, Dragos Costea et al.

Semantic segmentation is a crucial task for robot navigation and safety. However, current supervised methods require a large amount of pixelwise annotations to yield accurate results. Labeling is a tedious and time consuming process that has hampered progress in low altitude UAV applications. This paper makes an important step towards automatic annotation by introducing SegProp, a novel iterative flow-based method, with a direct connection to spectral clustering in space and time, to propagate the semantic labels to frames that lack human annotations. The labels are further used in semi-supervised learning scenarios. Motivated by the lack of a large video aerial dataset, we also introduce Ruralscapes, a new dataset with high resolution (4K) images and manually-annotated dense labels every 50 frames - the largest of its kind, to the best of our knowledge. Our novel SegProp automatically annotates the remaining unlabeled 98% of frames with an accuracy exceeding 90% (F-measure), significantly outperforming other state-of-the-art label propagation methods. Moreover, when integrating other methods as modules inside SegProp's iterative label propagation loop, we achieve a significant boost over the baseline labels. Finally, we test SegProp in a full semi-supervised setting: we train several state-of-the-art deep neural networks on the SegProp-automatically-labeled training frames and test them on completely novel videos. We convincingly demonstrate, every time, a significant improvement over the supervised scenario.

Iulia Duta, Andrei Nicolicioiu, Marius Leordeanu

Graph Neural Networks are perfectly suited to capture latent interactions between various entities in the spatio-temporal domain (e.g. videos). However, when an explicit structure is not available, it is not obvious what atomic elements should be represented as nodes. Current works generally use pre-trained object detectors or fixed, predefined regions to extract graph nodes. Improving upon this, our proposed model learns nodes that dynamically attach to well-delimited salient regions, which are relevant for a higher-level task, without using any object-level supervision. Constructing these localized, adaptive nodes gives our model inductive bias towards object-centric representations and we show that it discovers regions that are well correlated with objects in the video. In extensive ablation studies and experiments on two challenging datasets, we show superior performance to previous graph neural networks models for video classification.

Marian Petrica, Radu D. Stochitoiu, Marius Leordeanu et al.

In this paper we propose a three stages analysis of the evolution of Covid19 in Romania. There are two main issues when it comes to pandemic prediction. The first one is the fact that the numbers reported of infected and recovered are unreliable, however the number of deaths is more accurate. The second issue is that there were many factors which affected the evolution of the pandemic. In this paper we propose an analysis in three stages. The first stage is based on the classical SIR model which we do using a neural network. This provides a first set of daily parameters. In the second stage we propose a refinement of the SIR model in which we separate the deceased into a distinct category. By using the first estimate and a grid search, we give a daily estimation of the parameters. The third stage is used to define a notion of turning points (local extremes) for the parameters. We call a regime the time between these points. We outline a general way based on time varying parameters of SIRD to make predictions.

Radu D. Stochiţoiu, Marian Petrica, Traian Rebedea et al.

Analysing and understanding the transmission and evolution of the COVID-19 pandemic is mandatory to be able to design the best social and medical policies, foresee their outcomes and deal with all the subsequent socio-economic effects. We address this important problem from a computational and machine learning perspective. More specifically, we want to statistically estimate all the relevant parameters for the new coronavirus COVID-19, such as the reproduction number, fatality rate or length of infectiousness period, based on Romanian patients, as well as be able to predict future outcomes. This endeavor is important, since it is well known that these factors vary across the globe, and might be dependent on many causes, including social, medical, age and genetic factors. We use a recently published improved version of SEIR, which is the classic, established model for infectious diseases. We want to infer all the parameters of the model, which govern the evolution of the pandemic in Romania, based on the only reliable, true measurement, which is the number of deaths. Once the model parameters are estimated, we are able to predict all the other relevant measures, such as the number of exposed and infectious people. To this end, we propose a self-supervised approach to train a deep convolutional network to guess the correct set of Modified-SEIR model parameters, given the observed number of daily fatalities. Then, we refine the solution with a stochastic coordinate descent approach. We compare our deep learning optimization scheme with the classic grid search approach and show great improvement in both computational time and prediction accuracy. We find an optimistic result in the case fatality rate for Romania which may be around 0.3% and we also demonstrate that our model is able to correctly predict the number of daily fatalities for up to three weeks in the future.

Florin Condrea, Victor-Andrei Ivan, Marius Leordeanu

Automatically detecting vital signs in videos, such as the estimation of heart and respiration rates, is a challenging research problem in computer vision with important applications in the medical field. One of the key difficulties in tackling this task is the lack of sufficient supervised training data, which severely limits the use of powerful deep neural networks. In this paper we address this limitation through a novel deep learning approach, in which a recurrent deep neural network is trained to detect vital signs in the infrared thermal domain from purely synthetic data. What is most surprising is that our novel method for synthetic training data generation is general, relatively simple and uses almost no prior medical domain knowledge. Moreover, our system, which is trained in a purely automatic manner and needs no human annotation, also learns to predict the respiration or heart intensity signal for each moment in time and to detect the region of interest that is most relevant for the given task, e.g. the nose area in the case of respiration. We test the effectiveness of our proposed system on the recent LCAS dataset and obtain state-of-the-art results.

Alina Marcu, Dragos Costea, Vlad Licaret et al.

Semantic segmentation is a crucial task for robot navigation and safety. However, it requires huge amounts of pixelwise annotations to yield accurate results. While recent progress in computer vision algorithms has been heavily boosted by large ground-level datasets, the labeling time has hampered progress in low altitude UAV applications, mostly due to the difficulty imposed by large object scales and pose variations. Motivated by the lack of a large video aerial dataset, we introduce a new one, with high resolution (4K) images and manually-annotated dense labels every 50 frames. To help the video labeling process, we make an important step towards automatic annotation and propose SegProp, an iterative flow-based method with geometric constrains to propagate the semantic labels to frames that lack human annotations. This results in a dataset with more than 50k annotated frames - the largest of its kind, to the best of our knowledge. Our experiments show that SegProp surpasses current state-of-the-art label propagation methods by a significant margin. Furthermore, when training a semantic segmentation deep neural net using the automatically annotated frames, we obtain a compelling overall performance boost at test time of 16.8% mean F-measure over a baseline trained only with manually-labeled frames. Our Ruralscapes dataset, the label propagation code and a fast segmentation tool are available at our website: https://sites.google.com/site/aerialimageunderstanding/

Petru Soviany, Claudiu Ardei, Radu Tudor Ionescu et al.

Despite the significant advances in recent years, Generative Adversarial Networks (GANs) are still notoriously hard to train. In this paper, we propose three novel curriculum learning strategies for training GANs. All strategies are first based on ranking the training images by their difficulty scores, which are estimated by a state-of-the-art image difficulty predictor. Our first strategy is to divide images into gradually more difficult batches. Our second strategy introduces a novel curriculum loss function for the discriminator that takes into account the difficulty scores of the real images. Our third strategy is based on sampling from an evolving distribution, which favors the easier images during the initial training stages and gradually converges to a uniform distribution, in which samples are equally likely, regardless of difficulty. We compare our curriculum learning strategies with the classic training procedure on two tasks: image generation and image translation. Our experiments indicate that all strategies provide faster convergence and superior results. For example, our best curriculum learning strategy applied on spectrally normalized GANs (SNGANs) fooled human annotators in thinking that generated CIFAR-like images are real in 25.0% of the presented cases, while the SNGANs trained using the classic procedure fooled the annotators in only 18.4% cases. Similarly, in image translation, the human annotators preferred the images produced by the Cycle-consistent GAN (CycleGAN) trained using curriculum learning in 40.5% cases and those produced by CycleGAN based on classic training in only 19.8% cases, 39.7% cases being labeled as ties.

Iulia Paraicu, Marius Leordeanu

When driving, people make decisions based on current traffic as well as their desired route. They have a mental map of known routes and are often able to navigate without needing directions. Current self-driving models improve their performances when using additional GPS information. Here we aim to push forward self-driving research and perform route planning even in the absence of GPS. Our system learns to predict in real-time vehicle's current location and future trajectory, as a function of time, on a known map, given only the raw video stream and the intended destination. The GPS signal is available only at training time, with training data annotation being fully automatic. Different from other published models, we predict the vehicle's trajectory for up to seven seconds ahead, from which complete steering, speed and acceleration information can be derived for the entire time span. Trajectories capture navigational information on multiple levels, from instant steering commands that depend on present traffic and obstacles ahead, to longer-term navigation decisions, towards a specific destination. We collect our dataset with a regular car and a smartphone that records video and GPS streams. The GPS data is used to derive ground-truth supervision labels and create an analytical representation of the traversed map. In tests, our system outperforms published methods on visual localization and steering and gives accurate navigation assistance between any two known locations.

Emanuela Haller, Adina Magda Florea, Marius Leordeanu

We address the challenging task of foreground object discovery and segmentation in video. We introduce an efficient solution, suitable for both unsupervised and supervised scenarios, based on a spacetime graph representation of the video sequence. We ensure a fine grained representation with one-to-one correspondences between graph nodes and video pixels. We formulate the task as a spectral clustering problem by exploiting the spatio-temporal consistency between the scene elements in terms of motion and appearance. Graph nodes that belong to the main object of interest should form a strong cluster, as they are linked through long range optical flow chains and have similar motion and appearance features along those chains. On one hand, the optimization problem aims to maximize the segmentation clustering score based on the motion structure through space and time. On the other hand, the segmentation should be consistent with respect to node features. Our approach leads to a graph formulation in which the segmentation solution becomes the principal eigenvector of a novel Feature-Motion matrix. While the actual matrix is not computed explicitly, the proposed algorithm efficiently computes, in a few iteration steps, the principal eigenvector that captures the segmentation of the main object in the video. The proposed algorithm, GO-VOS, produces a global optimum solution and, consequently, it does not depend on initialization. In practice, GO-VOS achieves state of the art results on three challenging datasets used in current literature: DAVIS, SegTrack and YouTube-Objects.

Elena Burceanu, Marius Leordeanu

We formulate object segmentation in video as a graph partitioning problem in space and time, in which nodes are pixels and their relations form local neighborhoods. We claim that the strongest cluster in this pixel-level graph represents the salient object segmentation. We compute the main cluster using a novel and fast 3D filtering technique that finds the spectral clustering solution, namely the principal eigenvector of the graph's adjacency matrix, without building the matrix explicitly - which would be intractable. Our method is based on the power iteration for finding the principal eigenvector of a matrix, which we prove is equivalent to performing a specific set of 3D convolutions in the space-time feature volume. This allows us to avoid creating the matrix and have a fast parallel implementation on GPU. We show that our method is much faster than classical power iteration applied directly on the adjacency matrix. Different from other works, ours is dedicated to preserving object consistency in space and time at the level of pixels. For that, it requires powerful pixel-wise features at the frame level. This makes it perfectly suitable for incorporating the output of a backbone network or other methods and fast-improving over their solution without supervision. In experiments, we obtain consistent improvement, with the same set of hyper-parameters, over the top state of the art methods on DAVIS-2016 dataset, both in unsupervised and semi-supervised tasks. We also achieve top results on the well-known SegTrackv2 dataset.

Andretti Naiden, Vlad Paunescu, Gyeongmo Kim et al.

We propose Shift R-CNN, a hybrid model for monocular 3D object detection, which combines deep learning with the power of geometry. We adapt a Faster R-CNN network for regressing initial 2D and 3D object properties and combine it with a least squares solution for the inverse 2D to 3D geometric mapping problem, using the camera projection matrix. The closed-form solution of the mathematical system, along with the initial output of the adapted Faster R-CNN are then passed through a final ShiftNet network that refines the result using our newly proposed Volume Displacement Loss. Our novel, geometrically constrained deep learning approach to monocular 3D object detection obtains top results on KITTI 3D Object Detection Benchmark, being the best among all monocular methods that do not use any pre-trained network for depth estimation.

Andrei Nicolicioiu, Iulia Duta, Marius Leordeanu

Learning in the space-time domain remains a very challenging problem in machine learning and computer vision. Current computational models for understanding spatio-temporal visual data are heavily rooted in the classical single-image based paradigm. It is not yet well understood how to integrate information in space and time into a single, general model. We propose a neural graph model, recurrent in space and time, suitable for capturing both the local appearance and the complex higher-level interactions of different entities and objects within the changing world scene. Nodes and edges in our graph have dedicated neural networks for processing information. Nodes operate over features extracted from local parts in space and time and previous memory states. Edges process messages between connected nodes at different locations and spatial scales or between past and present time. Messages are passed iteratively in order to transmit information globally and establish long range interactions. Our model is general and could learn to recognize a variety of high level spatio-temporal concepts and be applied to different learning tasks. We demonstrate, through extensive experiments and ablation studies, that our model outperforms strong baselines and top published methods on recognizing complex activities in video. Moreover, we obtain state-of-the-art performance on the challenging Something-Something human-object interaction dataset.

Ioana Croitoru, Simion-Vlad Bogolin, Marius Leordeanu

Unsupervised learning poses one of the most difficult challenges in computer vision today. The task has an immense practical value with many applications in artificial intelligence and emerging technologies, as large quantities of unlabeled videos can be collected at relatively low cost. In this paper, we address the unsupervised learning problem in the context of detecting the main foreground objects in single images. We train a student deep network to predict the output of a teacher pathway that performs unsupervised object discovery in videos or large image collections. Our approach is different from published methods on unsupervised object discovery. We move the unsupervised learning phase during training time, then at test time we apply the standard feed-forward processing along the student pathway. This strategy has the benefit of allowing increased generalization possibilities during training, while remaining fast at testing. Our unsupervised learning algorithm can run over several generations of student-teacher training. Thus, a group of student networks trained in the first generation collectively create the teacher at the next generation. In experiments our method achieves top results on three current datasets for object discovery in video, unsupervised image segmentation and saliency detection. At test time the proposed system is fast, being one to two orders of magnitude faster than published unsupervised methods.

Iulia Duta, Andrei Liviu Nicolicioiu, Simion-Vlad Bogolin et al.

Describing visual data into natural language is a very challenging task, at the intersection of computer vision, natural language processing and machine learning. Language goes well beyond the description of physical objects and their interactions and can convey the same abstract idea in many ways. It is both about content at the highest semantic level as well as about fluent form. Here we propose an approach to describe videos in natural language by reaching a consensus among multiple encoder-decoder networks. Finding such a consensual linguistic description, which shares common properties with a larger group, has a better chance to convey the correct meaning. We propose and train several network architectures and use different types of image, audio and video features. Each model produces its own description of the input video and the best one is chosen through an efficient, two-phase consensus process. We demonstrate the strength of our approach by obtaining state of the art results on the challenging MSR-VTT dataset.

Elena Burceanu, Marius Leordeanu

Object tracking is an essential problem in computer vision that has been researched for several decades. One of the main challenges in tracking is to adapt to object appearance changes over time and avoiding drifting to background clutter. We address this challenge by proposing a deep neural network composed of different parts, which functions as a society of tracking parts. They work in conjunction according to a certain policy and learn from each other in a robust manner, using co-occurrence constraints that ensure robust inference and learning. From a structural point of view, our network is composed of two main pathways. One pathway is more conservative. It carefully monitors a large set of simple tracker parts learned as linear filters over deep feature activation maps. It assigns the parts different roles. It promotes the reliable ones and removes the inconsistent ones. We learn these filters simultaneously in an efficient way, with a single closed-form formulation, for which we propose novel theoretical properties. The second pathway is more progressive. It is learned completely online and thus it is able to better model object appearance changes. In order to adapt in a robust manner, it is learned only on highly confident frames, which are decided using co-occurrences with the first pathway. Thus, our system has the full benefit of two main approaches in tracking. The larger set of simpler filter parts offers robustness, while the full deep network learned online provides adaptability to change. As shown in the experimental section, our approach achieves state of the art performance on the challenging VOT17 benchmark, outperforming the published methods both on the general EAO metric and in the number of fails, by a significant margin.

Alina Marcu, Dragos Costea, Emil Slusanschi et al.

Semantic segmentation and vision-based geolocalization in aerial images are challenging tasks in computer vision. Due to the advent of deep convolutional nets and the availability of relatively low cost UAVs, they are currently generating a growing attention in the field. We propose a novel multi-task multi-stage neural network that is able to handle the two problems at the same time, in a single forward pass. The first stage of our network predicts pixelwise class labels, while the second stage provides a precise location using two branches. One branch uses a regression network, while the other is used to predict a location map trained as a segmentation task. From a structural point of view, our architecture uses encoder-decoder modules at each stage, having the same encoder structure re-used. Furthermore, its size is limited to be tractable on an embedded GPU. We achieve commercial GPS-level localization accuracy from satellite images with spatial resolution of 1 square meter per pixel in a city-wide area of interest. On the task of semantic segmentation, we obtain state-of-the-art results on two challenging datasets, the Inria Aerial Image Labeling dataset and Massachusetts Buildings.

Elena Burceanu, Marius Leordeanu

Object tracking is an essential task in computer vision that has been studied since the early days of the field. Being able to follow objects that undergo different transformations in the video sequence, including changes in scale, illumination, shape and occlusions, makes the problem extremely difficult. One of the real challenges is to keep track of the changes in objects appearance and not drift towards the background clutter. Different from previous approaches, we obtain robustness against background with a tracker model that is composed of many different parts. They are classifiers that respond at different scales and locations. The tracker system functions as a society of parts, each having its own role and level of credibility. Reliable classifiers decide the tracker's next move, while newcomers are first monitored before gaining the necessary level of reliability to participate in the decision process. Some parts that loose their consistency are rejected, while others that show consistency for a sufficiently long time are promoted to permanent roles. The tracker system, as a whole, could also go through different phases, from the usual, normal functioning to states of weak agreement and even crisis. The tracker system has different governing rules in each state. What truly distinguishes our work from others is not necessarily the strength of individual tracking parts, but the way in which they work together and build a strong and robust organization. We also propose an efficient way to learn simultaneously many tracking parts, with a single closed-form formulation. We obtain a fast and robust tracker with state of the art performance on the challenging OTB50 dataset.