Mihai Pîrvu

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 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.

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.

Mihai Cristian Pîrvu

We propose a method at the intersection of Computer Vision and Computer Graphics fields, which automatically generates RGBD images using neural networks, based on previously seen and synchronized video, depth and pose signals. Since the models must be able to reconstruct both texture (RGB) and structure (Depth), it creates an implicit representation of the scene, as opposed to explicit ones, such as meshes or point clouds. The process can be thought of as neural rendering, where we obtain a function f : Pose -> RGBD, which we can use to navigate through the generated scene, similarly to graphics simulations. We introduce two new datasets, one based on synthetic data with full ground truth information, while the other one being recorded from a drone flight in an university campus, using only video and GPS signals. Finally, we propose a fully unsupervised method of generating datasets from videos alone, in order to train the Pose2RGBD networks. Code and datasets are available at:: https://gitlab.com/mihaicristianpirvu/pose2rgbd.

Mihai Cristian Pîrvu, Alexandra Anghel

E-commerce web applications are almost ubiquitous in our day to day life, however as useful as they are, most of them have little to no adaptation to user needs, which in turn can cause both lower conversion rates as well as unsatisfied customers. We propose a machine learning system which learns the user behaviour from multiple previous sessions and predicts useful metrics for the current session. In turn, these metrics can be used by the applications to customize and better target the customer, which can mean anything from offering better offers of specific products, targeted notifications or placing smart ads. The data used for the learning algorithm is extracted from Google Analytics Enhanced E-commerce, which is enabled by most e-commerce websites and thus the system can be used by any such merchant. In order to learn the user patterns, only its behaviour features were used, which don't include names, gender or any other personal information that could identify the user. The learning model that was used is a double recurrent neural network which learns both intra-session and inter-session features. The model predicts for each session a probability score for each of the defined target classes.

Mihai Cristian Pîrvu, Alexandra Anghel, Ciprian Borodescu et al.

Predicting user behaviour on a website is a difficult task, which requires the integration of multiple sources of information, such as geo-location, user profile or web surfing history. In this paper we tackle the problem of predicting the user intent, based on the queries that were used to access a certain webpage. We make no additional assumptions, such as domain detection, device used or location, and only use the word information embedded in the given query. In order to build competitive classifiers, we label a small fraction of the EDI query intent prediction dataset \cite{edi-challenge-dataset}, which is used as ground truth. Then, using various rule-based approaches, we automatically label the rest of the dataset, train the classifiers and evaluate the quality of the automatic labeling on the ground truth dataset. We used both recurrent and convolutional networks as the models, while representing the words in the query with multiple embedding methods.