INFuture2017 - Integrating ICT in Society
International audience
Evaluating the Acquisition of Semantic Knowledge from Cross-situational Learning in Artificial Neural Networks
When learning their native language, children acquire the meanings of words and sentences from highly ambiguous input without much explicit supervision. One possible learning mechanism is cross-situational learning, which has been successfully tested in laboratory experiments with children. Here we use Artificial Neural Networks to test if this mec...
Compiling pattern matching to in-place modifications / Compiling pattern matching to in-place modifications: (short pape...
Algebraic data types and pattern matching are popular tools to build programs manipulating complex datastructures in a safe yet efficient manner. On top of its safety advantages, compilation techniques can turn pattern matching into highly efficient deconstruction code for immutable use cases. Conversely, high-performance datastructures and languag...
Data Abstraction: A General Framework to Handle Program Verification of Data Structures
Proving properties on programs accessing data structures such as arrays often requires universally quantified invariants, e.g., "all elements below index i are nonzero". In this article, we propose a general data abstraction scheme operating on Horn formulas, into which we recast previously published abstractions. We show that our instantiation sch...
Just Ask: Learning to Answer Questions from Millions of Narrated Videos
Recent methods for visual question answering rely on large-scale annotated datasets. Manual annotation of questions and answers for videos, however, is tedious, expensive and prevents scalability. In this work, we propose to avoid manual annotation and generate a large-scale training dataset for video question answering making use of automatic cros...
Execution Trace Analysis for a Precise Understanding of Latency Violations
Despite the amount of proposed works for the verification of diverse model properties, understanding the root cause of latency requirements violation in execution traces is still an open-issue especially for complex HW/SW system-level designs: is it due to an unfavorable real-time scheduling, to contentions on buses, to the characteristics of funct...
Enhancing Speech Privacy with Slicing
Privacy preservation calls for speech anonymization methods which hide the speaker's identity while minimizing the impact on downstream tasks such as automatic speech recognition (ASR) training or decoding. In the recent VoicePrivacy 2020 Challenge, several anonymization methods have been proposed to transform speech utterances in a way that preser...
A Neural Approach for Detecting Morphological Analogies
International audience
Transformer versus LSTM Language Models Trained on Uncertain ASR Hypotheses in Limited Data Scenarios
In several ASR use cases, training and adaptation of domain-specific LMs can only rely on a small amount of manually verified text transcriptions and sometimes a limited amount of in-domain speech. Training of LSTM LMs in such limited data scenarios can benefit from alternate uncertain ASR hypotheses, as observed in our recent work. In this paper, ...
GPT-2's activations predict the degree of semantic comprehension in the human brain
Language transformers, like GPT-2, have demonstrated remarkable abilities to process text, and now constitute the backbone of deep translation, summarization and dialogue algorithms. However, whether these models encode information that relates to human comprehension remains controversial. Here, we show that the representations of GPT-2 not only ma...
