Sökning: "Bottom-up processing"

Visar resultat 1 - 5 av 18 uppsatser innehållade orden Bottom-up processing.

1. 1. Approximating Reasoning with Transformer Language Models

   Master-uppsats, Göteborgs universitet/Institutionen för data- och informationsteknik

   Författare :Daniel Enström; Viktor Kjellberg; [2023-10-19]
   Nyckelord :natural language processing; nlp; automated reasoning; logic; inference; machine learning; transformers; language models; BERT; BART;

   Sammanfattning : We conduct experiments with BART, a generative language-model architecture, to investigate its capabilities for approximating reasoning by learning from data. For this we use the SimpleLogic dataset, a dataset of satisfiability problems in propositional logic originally created by Zhang et al. (2022). LÄS MER

3. 2. Trilingual spoken word recognition : Interlingual competition from one or two non-target languages in a sentence context

   Magister-uppsats, Stockholms universitet/Centrum för tvåspråkighetsforskning

   Författare :Yulia Kashevarova; [2023]
   Nyckelord :trilingual speech processing; cross-linguistic competition; sentence context; BLINCS; BIA ;

   Sammanfattning : Persistent non-target language co-activation in spoken and visual language comprehension has been found both at the word-level and at the level of a sentence, although in the latter case, sentence bias has been observed to modulate the co-activation which can create lexical competition. In the case of trilingual speakers, both non-target languages may potentially compete with the third language (L3). LÄS MER

4. 3. Human pose estimation in low-resolution images

   Master-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)

   Författare :Hugo Nilsson; [2022]
   Nyckelord :Human Pose Estimation; Low-Resolution Images; Heatmap Decoding; Computer Vision; Deep Learning; Mänsklig Poseuppskattning; Lågupplösta Bilder; Färgdiagram Avkodning; Datorseende; Djupinlärning;

   Sammanfattning : This project explores the understudied, yet important, case of human pose estimation in low-resolution images. This is done in the use-case of images with football players of known scale in the image. Human pose estimation can mainly be done in two different ways, the bottom-up method and the top-down method. LÄS MER

5. 4. Deriving an Natural Language Processing inference Cost Model with Greenhouse Gas Accounting : Towards a sustainable usage of Machine Learning

   Uppsats för yrkesexamina på avancerad nivå, KTH/Skolan för elektroteknik och datavetenskap (EECS)

   Författare :Tom Axberg; [2022]
   Nyckelord :Machine Learning; Inference; Inferencing; Natural Language Processing; Pre-trained Language Model; Greenhouse Gas; Software Energy Measurement; Floating Point Operations; Green-AI; Green-IT; OPEX; Maskininlärning; Slutledning; Naturlig Språkbehandling; Förutbildad språkmodell; Växthusgas; Energimätning av Mjukvara; Green-AI; Green-IT. OPEX;

   Sammanfattning : The interest in using State-Of-The-Art (SOTA) Pre-Trained Language Model (PLM) in product development is growing. The fact that developers can use PLM has changed the way to build reliable models, and it is the go-to method for many companies and organizations. LÄS MER

7. 5. Film clubbing : collective rituals for processing eco-anxiety

   Master-uppsats, SLU/Dept. of Urban and Rural Development

   Författare :Felicitas Paul; [2022]
   Nyckelord :Film clubs; eco-anxiety; processing; emotions; interaction rituals; eudaimonic entertainment;

   Sammanfattning : Research shows that an increasing number of people suffer from the emotional consequences arising from the confrontation with climate crisis (CC), often referred to as eco-anxiety. Due to a lack of opportunities to express and process these emotions, a common response is to withdraw from talking about the CC leading to a constructed silence around the topic. LÄS MER