Tootfinder

Opt-in global Mastodon full text search. Join the index!

No exact results. Similar results found.
@arXiv_csCL_bot@mastoxiv.page
2025-09-17 09:16:00

MAGIC-Enhanced Keyword Prompting for Zero-Shot Audio Captioning with CLIP Models
Vijay Govindarajan, Pratik Patel, Sahil Tripathi, Md Azizul Hoque, Gautam Siddharth Kashyap
arxiv.org/abs/2509.12591

MAGIC-Enhanced Keyword Prompting for Zero-Shot Audio Captioning with CLIP Models
Automated Audio Captioning (AAC) generates captions for audio clips but faces challenges due to limited datasets compared to image captioning. To overcome this, we propose the zero-shot AAC system that leverages pre-trained models, eliminating the need for extensive training. Our approach uses a pre-trained audio CLIP model to extract auditory features and generate a structured prompt, which guides a Large Language Model (LLM) in caption generation. Unlike traditional greedy decoding, our metho…

@pbloem@sigmoid.social
2025-07-18 09:25:22

Now out in #TMLR:
🍇 GRAPES: Learning to Sample Graphs for Scalable Graph Neural Networks 🍇
There's lots of work on sampling subgraphs for GNNs, but relatively little on making this sampling process _adaptive_. That is, learning to select the data from the graph that is relevant for your task.
We introduce an RL-based and a GFLowNet-based sampler and show that the approach perf…

A diagram of the GRAPES pipeline. It shows a subgraph being sampled in two steps and being fed to a GNN, with a blue line showing the learning signal. The caption reads Figure 1: Overview of GRAPES. First, GRAPES processes a target node (green) by computing node inclusion probabilities on its 1-hop neighbors (shown by node color shade) with a sampling GNN. Given these probabilities, GRAPES samples k nodes. Then, GRAPES repeats this process over nodes in the 2-hop neighborhood. We pass the sampl…
A results table for node classification on heterophilious graphs. Table 2: F1-scores (%) for different sampling methods trained on heterophilous graphs for a batch size of 256, and a sample size of 256 per layer. We report the mean and standard deviation over 10 runs. The best values among the sampling baselines (all except GAS) are in bold, and the second best are underlined. MC stands for multi-class and ML stands for multi-label classification. OOM indicates out of memory.
Performance of samples vs sampling size showing that GRAPES generally performs well across sample sizes, while other samplers often show more variance across sample sizes. The caption reads Figure 4: Comparative analysis of classification accuracy across different sampling sizes for sampling baseline and GRAPES. We repeated each experiment five times: The shaded regions show the 95% confidence intervals.
A diagrammatic illustration of a graph classification task used in one of the theorems. The caption reads Figure 9: An example of a graph for Theorem 1 with eight nodes. Red edges belong to E1, features xi and labels yi are shown beside every node. For nodes v1 and v2 we show the edge e12 as an example. As shown, the label of each node is the second feature of its neighbor, where a red edge connects them. The edge homophily ratio is h=12/28 = 0.43.
@radioeinsmusicbot@mastodonapp.uk
2025-09-15 04:24:44

🇺🇦 Auf radioeins läuft...
Cupidon feat. Milaa:
🎵 Feel It
#NowPlaying #Cupidon #Milaa
subreachers.bandcamp.com/track
open.spotify.com/track/5qKEArG

@qurlyjoe@mstdn.social
2025-08-17 01:40:12

#caturday

Meme. A cat with its mouth wide open, showing lots of teeth, is sitting in a box which it has apparently been chewing the crap out of. Caption: I survived. Suck it Schrödinger.
@matthiasott@mastodon.social
2025-09-18 13:46:52

Had an amazing time speaking about Web Design Engineering at @… Freiburg last week! 🎉 It was an honour to be invited and to meet so many wonderful people and good friends there – a truly smashing experience! Thank you, everyone! 🤗💚🎈
📸 Photos by @…

Matthias on stage at Smashing Conf Freiburg, talking to the audience, with a monitor behind me displaying live captions.
Me on stage, viewed from afar with a truckload of modern CSS properties and functions on the screen behind me.
Vitaly Friedman and I talking on a red sofa during the Q&A after the talk.
@paulomalley@c.im
2025-09-16 22:18:01

I think my most-sent email of all time is "Does Tuesday at 2 pm work for you?" followed closely by "No worries, how about Wednesday?" 🫠
But what if you could just... not do that anymore?
There’s a scheduling tool hiding in plain sight in Gmail that ends all this nonsense. You offer times, they click once, and poof—it's in the calendar.
I filmed a quick guide on it because this needs to be public knowledge:

YouTube Thumbnail image featuring the Google logo and the caption "Schedule Meetings 10x faster!!!"
@seav@en.osm.town
2025-08-18 13:19:46

I finished watching this #Netflix documentary about the #Antwerp 🇧🇪 diamond heist in 2003 and while the film itself is your typical interesting and informative Netflix-style docu, I am again peeved that #OpenStreetMap

Frame from the 2025 documentary Stolen: Heist of the Century showing a desaturated basemap of central Antwerp, Belgium, centered around the diamond district with a red pin at the center labeled “Diamond Center” and with an accompanying photo of the building’s faade. A subtitle caption says, “under the cell tower of Diamond Center.”
@arXiv_csCV_bot@mastoxiv.page
2025-09-17 14:04:13

Replaced article(s) found for cs.CV. arxiv.org/list/cs.CV/new
[4/4]:
- Evaluating the Robustness of Open-Source Vision-Language Models to Domain Shift in Object Captioning
Federico Tavella, Amber Drinkwater, Angelo Cangelosi

@NFL@darktundra.xyz
2025-10-12 17:24:26

The unique way Patriots coach Mike Vrabel selects weekly game captains: 'It's a dope thing' espn.com/nfl/story/_/id/465577…

How Patriots coach Mike Vrabel selects weekly game captains - ESPN
This week, Mike Reiss highlights stats to know, rookie returners and where the lineman had dinner this week.

@arXiv_csAI_bot@mastoxiv.page
2025-09-17 13:53:20

Replaced article(s) found for cs.AI. arxiv.org/list/cs.AI/new
[4/5]:
- Evaluating the Robustness of Open-Source Vision-Language Models to Domain Shift in Object Captioning
Federico Tavella, Amber Drinkwater, Angelo Cangelosi