Future prediction with deep learning
Abstract: Future prediction is a fundamental principle of intelligence in which the future state of an environment is predicted given its past states. Accurate future prediction is relevant for applications that require safety in planning such as autonomous driving, robot navigation, or surveillance systems. The complexity of the task stems from the integration of multiple sources of information such as the past states of dynamic agents, the interactions among agents and the semantic constraints of the environment. Moreover, as the future is uncertain to a large extent, modeling the uncertainty and multimodality of the future states is of great relevance. The first part of the thesis proposes a sampling-fitting framework with a fully probabilistic output allowing the system to predict multiple future modes along with the uncertainty of each mode. It achieves the best trade-off between ensuring diversity of the prediction and matching the underlying true distribution of the future. The framework is applied to different settings of the task including the bird’s eye and egocentric views. For the latter, the framework is further extended to a multi-stages pipeline in which a general prior is generated, propagated to the future and finally refined to obtain the desired output. On both synthetic and large-scale real datasets, our framework triggers good estimates of multimodal distributions and avoids mode collapse. The second part of the thesis conducts an in-depth analysis of the task to uncover new challenges and seek better models. First, our analysis shows that easy scenarios dominate existing real datasets and the most critical ones are much less frequent, harder to learn and usually ignored by existing models. Therefore, we propose to reshape the learned feature space of existing models by pushing challenging scenarios closer to each other that triggers sharing relevant information and yields consequently better results. Second, given the black-box nature of existing models, it remains unclear which features are used to make a prediction. Therefore, we propose a method to quantify the contribution of different cues on the prediction. On common benchmarks, our analysis shows that existing methods are unable to reason about the interaction features between agents and the past state of the target agent is the only feature used for predicting its future
- Location
-
Deutsche Nationalbibliothek Frankfurt am Main
- Extent
-
Online-Ressource
- Language
-
Englisch
- Notes
-
Universität Freiburg, Dissertation, 2022
- Keyword
-
Maschinelles Lernen
Computersimulation
Autonomes Fahrzeug
Neuronales Netz
Maschinelles Sehen
Deep learning
- Event
-
Veröffentlichung
- (where)
-
Freiburg
- (who)
-
Universität
- (when)
-
2022
- Creator
- DOI
-
10.6094/UNIFR/230631
- URN
-
urn:nbn:de:bsz:25-freidok-2306317
- Rights
-
Open Access; Der Zugriff auf das Objekt ist unbeschränkt möglich.
- Last update
-
25.03.2025, 1:43 PM CET
Data provider
Deutsche Nationalbibliothek. If you have any questions about the object, please contact the data provider.
Associated
Time of origin
- 2022
Other Objects (12)
A deep learning approach for complex microstructure inference
Stable deep reinforcement learning
Learning navigation policies with deep reinforcement learning
Patient groups in rheumatoid arthritis identified by deep learning respond differently to biologic or targeted synthetic DMARDs
Scalable 3D deep learning: methods and applications
Towards efficient deep learning for computer vision
Deep learning on biomedical data with Boltzmann machines
Tool recommender system in Galaxy using deep learning
One-shot battery degradation trajectory prediction with deep learning
A deep learning approach for location independent throughput prediction
Refuting student teachers’ misconceptions about multimedia learning
DIGGER-Bac: prediction of seed regions for high-fidelity construction of synthetic small RNAs in bacteria
A deep learning approach for complex microstructure inference
Stable deep reinforcement learning
Learning navigation policies with deep reinforcement learning
Patient groups in rheumatoid arthritis identified by deep learning respond differently to biologic or targeted synthetic DMARDs
Scalable 3D deep learning: methods and applications
Towards efficient deep learning for computer vision
Deep learning on biomedical data with Boltzmann machines
Tool recommender system in Galaxy using deep learning
One-shot battery degradation trajectory prediction with deep learning
A deep learning approach for location independent throughput prediction
Refuting student teachers’ misconceptions about multimedia learning
DIGGER-Bac: prediction of seed regions for high-fidelity construction of synthetic small RNAs in bacteria
A deep learning approach for complex microstructure inference
Stable deep reinforcement learning
Learning navigation policies with deep reinforcement learning
Patient groups in rheumatoid arthritis identified by deep learning respond differently to biologic or targeted synthetic DMARDs
Scalable 3D deep learning: methods and applications
Towards efficient deep learning for computer vision
Deep learning on biomedical data with Boltzmann machines
Tool recommender system in Galaxy using deep learning
One-shot battery degradation trajectory prediction with deep learning
A deep learning approach for location independent throughput prediction
Refuting student teachers’ misconceptions about multimedia learning