Simulation-guided Beam Search for Neural Combinatorial Optimization [PDF]
Neural Information Processing Systems, 2022 Neural approaches for combinatorial optimization (CO) equip a learning mechanism to discover powerful heuristics for solving complex real-world problems.
Jinho Choo +6 more
semanticscholar +1 more source
Don’t Say What You Don’t Know: Improving the Consistency of Abstractive Summarization by Constraining Beam Search [PDF]
IEEE Games Entertainment Media Conference, 2022 ive summarization systems today produce fluent and relevant output, but often “hallucinate” statements not supported by the source text. We analyze the connection between hallucinations and training data, and find evidence that models hallucinate because
Daniel King +5 more
semanticscholar +1 more source
On the Feasibility of Out-of-Band Spatial Channel Information for Millimeter-Wave Beam Search [PDF]
IEEE Transactions on Antennas and Propagation, 2022 Prolonged beam alignment is the main source of overhead in mobile wireless communications at millimeter-wave (mm-wave) frequencies due to narrow beams following the requirement of high antenna gains. Out-of-band spatial information may be used in initial
Peize Zhang +4 more
semanticscholar +1 more source
Beyond Greedy Search: Tracking by Multi-Agent Reinforcement Learning-Based Beam Search [PDF]
IEEE Transactions on Image Processing, 2022 To track the target in a video, current visual trackers usually adopt greedy search for target object localization in each frame, that is, the candidate region with the maximum response score will be selected as the tracking result of each frame. However,
Xiao Wang +4 more
semanticscholar +1 more source
Improved Beam Search for Hallucination Mitigation in Abstractive Summarization [PDF]
arXiv.org, 2022 Advancement in large pretrained language models has significantly improved their performance for conditional language generation tasks including summarization albeit with hallucinations.
A. Sridhar, Erik M. Visser
semanticscholar +1 more source
Beam Search Algorithm for Anti-Collision Trajectory Planning for Many-to-Many Encounter Situations with Autonomous Surface Vehicles. [PDF]
Sensors (Basel), 2020 A single anti-collision trajectory generation problem for an “own” vessel only is significantly different from the challenge of generating a whole set of safe trajectories for multi-surface vehicle encounter situations in the open sea.
Koszelew J +5 more
europepmc +2 more sources
Streaming parallel transducer beam search with fast-slow cascaded encoders [PDF]
Interspeech, 2022 Streaming ASR with strict latency constraints is required in many speech recognition applications. In order to achieve the required latency, streaming ASR models sacrifice accuracy compared to non-streaming ASR models due to lack of future input context.
Jay Mahadeokar +7 more
semanticscholar +1 more source
Machine Translation Decoding beyond Beam Search [PDF]
Conference on Empirical Methods in Natural Language Processing, 2021 Beam search is the go-to method for decoding auto-regressive machine translation models. While it yields consistent improvements in terms of BLEU, it is only concerned with finding outputs with high model likelihood, and is thus agnostic to whatever end ...
Rémi Leblond +7 more
semanticscholar +1 more source
Decoding Methods in Neural Language Generation: A Survey
Information, 2021 Neural encoder-decoder models for language generation can be trained to predict words directly from linguistic or non-linguistic inputs. When generating with these so-called end-to-end models, however, the NLG system needs an additional decoding ...
Sina Zarrieß +2 more
doaj +1 more source
Frontiers in Applied Mathematics and Statistics, 2023 Vehicle routing problems are a class of NP-hard combinatorial optimization problems which attract a lot of attention, as they have many practical applications.
Jorin Dornemann
doaj +1 more source