Photosynthesis and respiration rates of (A) <i>Acropora</i> sp. and (B) <i>P. verrucosa</i> for the four groups. by Malte Ostendarp (13217973)

Experimental environment and parameters. by Haisong Xu (141729)

“…First, an efficient channel attention mechanism (ECA) is inserted in the layer before the SPPF in the backbone network, which realizes efficient computation of channel attention and reduces redundant computation while decreasing the model complication. Second, using the Content-Aware ReAssembly of Features (CARAFE) module instead of the original nearest-neighbor up-sampling module achieves light weighting while allowing for better aggregation of contextual information within a larger sensory field, which effectively improves the diversity and effectiveness of the model. …”

Results of ablation experiments. by Haisong Xu (141729)

“…First, an efficient channel attention mechanism (ECA) is inserted in the layer before the SPPF in the backbone network, which realizes efficient computation of channel attention and reduces redundant computation while decreasing the model complication. Second, using the Content-Aware ReAssembly of Features (CARAFE) module instead of the original nearest-neighbor up-sampling module achieves light weighting while allowing for better aggregation of contextual information within a larger sensory field, which effectively improves the diversity and effectiveness of the model. …”

ECA structural model diagram. by Haisong Xu (141729)

“…First, an efficient channel attention mechanism (ECA) is inserted in the layer before the SPPF in the backbone network, which realizes efficient computation of channel attention and reduces redundant computation while decreasing the model complication. Second, using the Content-Aware ReAssembly of Features (CARAFE) module instead of the original nearest-neighbor up-sampling module achieves light weighting while allowing for better aggregation of contextual information within a larger sensory field, which effectively improves the diversity and effectiveness of the model. …”

YOLOv5s general structure diagram. by Haisong Xu (141729)

“…First, an efficient channel attention mechanism (ECA) is inserted in the layer before the SPPF in the backbone network, which realizes efficient computation of channel attention and reduces redundant computation while decreasing the model complication. Second, using the Content-Aware ReAssembly of Features (CARAFE) module instead of the original nearest-neighbor up-sampling module achieves light weighting while allowing for better aggregation of contextual information within a larger sensory field, which effectively improves the diversity and effectiveness of the model. …”

Heat maps for different models. by Haisong Xu (141729)

“…First, an efficient channel attention mechanism (ECA) is inserted in the layer before the SPPF in the backbone network, which realizes efficient computation of channel attention and reduces redundant computation while decreasing the model complication. Second, using the Content-Aware ReAssembly of Features (CARAFE) module instead of the original nearest-neighbor up-sampling module achieves light weighting while allowing for better aggregation of contextual information within a larger sensory field, which effectively improves the diversity and effectiveness of the model. …”

Defect category statistics. by Haisong Xu (141729)

“…First, an efficient channel attention mechanism (ECA) is inserted in the layer before the SPPF in the backbone network, which realizes efficient computation of channel attention and reduces redundant computation while decreasing the model complication. Second, using the Content-Aware ReAssembly of Features (CARAFE) module instead of the original nearest-neighbor up-sampling module achieves light weighting while allowing for better aggregation of contextual information within a larger sensory field, which effectively improves the diversity and effectiveness of the model. …”

CARAFE general structure. by Haisong Xu (141729)

“…First, an efficient channel attention mechanism (ECA) is inserted in the layer before the SPPF in the backbone network, which realizes efficient computation of channel attention and reduces redundant computation while decreasing the model complication. Second, using the Content-Aware ReAssembly of Features (CARAFE) module instead of the original nearest-neighbor up-sampling module achieves light weighting while allowing for better aggregation of contextual information within a larger sensory field, which effectively improves the diversity and effectiveness of the model. …”

ECN-YOLOv5s structure diagram. by Haisong Xu (141729)

“…First, an efficient channel attention mechanism (ECA) is inserted in the layer before the SPPF in the backbone network, which realizes efficient computation of channel attention and reduces redundant computation while decreasing the model complication. Second, using the Content-Aware ReAssembly of Features (CARAFE) module instead of the original nearest-neighbor up-sampling module achieves light weighting while allowing for better aggregation of contextual information within a larger sensory field, which effectively improves the diversity and effectiveness of the model. …”

Technical properties of tested UASs. by Marc Roswag (20598443)

“…However, despite the use of larger propellers for the multicopter UAS compared to previous studies, we observed a deterrence effect for all echolocation groups. …”

Theoretical framework. by Dinku Mechal (21273002)

“…Holding other variables constant, a one-unit increase in demand and supply side risks results in 19.1% and 13.1% lead to decrease in operational performance of manufacturing companies, respectively. …”

Supplementary file survey questioner annex. by Dinku Mechal (21273002)

“…Holding other variables constant, a one-unit increase in demand and supply side risks results in 19.1% and 13.1% lead to decrease in operational performance of manufacturing companies, respectively. …”

Engineering the Thermal and Energy-Storage Properties in Quantum Dots Using Dominant Faceting: The Case Study of Silicon by Pavel Galář (17095622)

“…Finally, the thermal oxidation of the synthesized QDs is completed at lower temperatures with increasing SFE, decreasing from 1065 to 970 °C and being > 150 °C lower in QDs than in the larger reference nanoparticles. …”

Differentiating leader-follower and avoidance interactions. by Juan Fernández-Gracia (721245)

Estimating RSA Variability across Treatment, Time, and Condition. by Hikari Otsuka (18605246)

Simplified scheme illustrating the inhibitory effect of nNOS on the PPP during the initiation of GSIS. by Ingmar Lundquist (46422)

The expression levels of putative VIP1 target genes that were upregulated by brushing in brushed <i>atb’’αβγδ</i> leaves. by Hyuk Sung Yoon (20208672)

Fig 5 - by Raghu Arghal (20598093)

Relationships between clay and quartz content, TOC, 2D-NMR solid OM, S₁, 2D-NMR light oil and fractal dimensions. by Changjun Ji (18285907)

“…This results in the formation of larger dissolution pores within the carbonated minerals [<a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0309346#pone.0309346.ref043" target="_blank">43</a>], leading to a decrease in the fractal dimensions D₁ and D₂. …”

Heirarchical occupancy best fit model eDNA detection probabilities. by Matthew M. Dougherty (3246573)