ashu_trv

4 ms·

1.
Search over the Visual World: off-the-shelf VLMs beat video embeddings(arxiv.org)
6 points by ashu_trv 2mo ago|2 comments
2.
Record and Replay, teach AI agents desktop workflows by showing them once(github.com)
1 points by ashu_trv 3mo ago|0 comments
3.
Show HN: Do Thought Streams Matter? A Benchmark of VLM Reasoning in Gemini 2.5(arxiv.org)
3 points by ashu_trv 6mo ago|0 comments
4.
Show HN: I packaged decade of video infra battle scars into tools for AI agents
7 points by ashu_trv 7mo ago|0 comments
5.
by ashu_trv 1y ago
I agree. I am big fan of o3, and GPT 5 is not the same, it's like going back to GPT-3 level stupidity. It doesn't care about context, feels super dumb.
6.
by ashu_trv 1y ago
What's really impressive here: Model-Agnostic Infrastructure: Any AI model—open-source, proprietary, LLM, or VLM—can instantly gain real-time vision capabilities. No more waiting for Google to open their doors. Immediate Availability:
7.
Live video feed for every multimodal model not just Gemini(videodb.io)
7 points by ashu_trv 1y ago|2 comments
8.
Show HN: VideoDB – 80 % fewer hallucinations on NFL game analysis(docs.videodb.io)
1 points by ashu_trv 1y ago|0 comments
9.
Lessons Learned Building MCP for Video Infrastructure Startup
2 points by ashu_trv 1y ago|0 comments
10.
by ashu_trv 1y ago
Yeah, we tried to solve the 1 and 3. 2nd is still an open problem. Can you share more about the MECE?
11.
Auto-Sync Your Docs, SDKs and Examples for LLMs and AI Agents(github.com)
6 points by ashu_trv 1y ago|3 comments
12.
by ashu_trv 1y ago
Keeping documentation and SDK updates aligned with evolving "LLM contexts" can quickly overwhelm dev teams. At VideoDB, we've built an open-source solution—Agent Toolkit—that automates syncing your docs, SDK versions, and exa
13.
Ask HN: Model to Analyse Financial Transactions
1 points by ashu_trv 2y ago|1 comments
14.
Underwhelming MCP vs Hype
4 points by ashu_trv 2y ago|10 comments
15.
Benchmarking vision-language models on OCR in dynamic video environments(arxiv.org)
142 points by ashu_trv 2y ago|58 comments
16.
Vision-Language Models vs. Traditional OCR in Video – New Benchmark(arxiv.org)
6 points by ashu_trv 2y ago|1 comments
17.
by ashu_trv 2y ago
A new benchmark study evaluates Vision-Language Models (Claude-3, Gemini-1.5, GPT-4o) against traditional OCR tools (EasyOCR, RapidOCR) for extracting text from videos. The findings show VLMs outperforming OCR in many cases but also highlig
18.
by ashu_trv 2y ago
Thanks! Added now.
19.
Show HN:Video is hard: until now(github.com)
4 points by ashu_trv 2y ago|4 comments
20.
by ashu_trv 2y ago
This open-source agent framework is like ChatGPT, but for videos. It simplifies complex video tasks like search, editing, compilation, and—best of all—generation. The results stream instantly. You can even extend the agents to suit your nee
21.
by ashu_trv 3y ago
It analyse the transcript, but there is no way to get back the video clip without building your own video infra. We at Videodb are solving the exact problem.
22.
by ashu_trv 3y ago
LLMs are great with text, but they don't help you consume or create video clips. Checkout PromptClip - Use natural language to describe the what you want. - Instantly get video clips with the help of LLMs like OpenAI or Claude. Few int
23.
by ashu_trv 3y ago
Yeah, VideoDB is the next-gen infrastructure for videos and actually less costly than current video infrastructure. But you can use any LLM for analysing the transcript.
24.
Show HN: Instantly create video clips from LLM prompts(github.com)
4 points by ashu_trv 3y ago|5 comments
25.
by ashu_trv 3y ago
Build custom GPT on your video data with StreamRAG in 2 mins.This search agent find relevant moments across hundreds of hours of content and return a video clip instantly. RAG applications are great with text, but with video they can'
26.
Show HN: GPT-Powered Video Retrieval and Streaming(github.com)
5 points by ashu_trv 3y ago|1 comments
27.
Show HN: Twitter bot generates interactive transcript of any audio/video(twitter.com)
7 points by ashu_trv 6y ago|2 comments
28.
by ashu_trv 6y ago
Made a twitter bot that lets you generate an interactive transcript(Spext Docs) of any audio/video in a tweet Here is an example - https://twitter.com/spext_it/status/1286130139290632192 Interactive transcrip
29.
by ashu_trv 7y ago
Very cool! How is it going to be beneficial over the native solutions provided by cloud providers itself?